New Relic vs Honeycomb

Agent access

1. ai-native userPoint an agent at llms.txt or agent-oriented docs

   weight 2 · round to Honeycomb
   New Relic—none0/10

   A direct probe confirms llms.txt returns 404, and there's no evidence of agent-oriented docs designed for AI consumption; the MCP server is a separate integration mechanism, not agent-readable documentation. Missing for 10: an llms.txt file, any agent-oriented/markdown docs format, or evidence of AI agents successfully consuming New Relic docs directly.

   • [probe] “PROBE llms.txt: HTTP 404 at https://docs.newrelic.com/llms.txt”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”

   proof ↗⚑ flag

   Honeycomb✓fullprobed9/10

   A direct probe confirms llms.txt is live at https://docs.honeycomb.io/llms.txt returning HTTP 200 with structured agent-readable docs, and Honeycomb also ships an official MCP server for agents to query docs/telemetry directly. missing for 10: no independent third-party report of an agent actually consuming llms.txt in practice.

   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.honeycomb.io/llms.txt # Honeycomb Docs ## Get Started - [Get Started: Overview](https://docs.hone…”
   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”

   proof ↗⚑ flag

2. ai-native userRun the product headlessly / in CI for automation

   weight 2 · round to New Relic
   New Relic~partialprobed6/10

   New Relic ships an official CLI (newrelic-cli) for deployment markers and entity search, plus a GraphQL-based NerdGraph API, both of which are inherently scriptable/headless and usable in CI pipelines (e.g. recording deployment markers on release). However, there is no explicit CI/CD integration guide, pipeline example, or documented automation workflow tying these together for AI-native/agentic use. Missing for 10: explicit CI/CD pipeline documentation or examples, evidence of headless operation being a first-class supported workflow, and independent confirmation of CLI use in automated pipelines.

   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [probe] “official CLI documented at https://github.com/newrelic/newrelic-cli”

   proof ↗⚑ flag

   Honeycomb~partialprobed5/10

   Honeycomb exposes a full API (and OpenAPI spec) for programmatically managing datasets, queries, triggers and SLOs, and OTel-based data ingestion is inherently headless-compatible, which supports scripted/CI automation. However there is no explicit CI/CD example, no documented CLI, and the probe shows the openapi.json spec itself 404s, so first-class 'headless in CI' support is only inferred rather than directly evidenced. Missing for 10: an explicit CI/CD pipeline example or GitHub Actions integration, a documented CLI tool, and a working OpenAPI spec download.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”
   • [claimed-docs] “If your application is already instrumented with OpenTelemetry, you can send OpenTelemetry Protocol (OTLP) data directly to Honeycomb.”

   proof ↗⚑ flag

3. ai-native userPlug MCP servers into this product so it can use their tools

   weight 3 · round drawn
   New Relic—none0/10

   Evidence shows New Relic ships an official MCP server so external AI tools can connect to New Relic's data (new-relic-docs-11, new-relic-docs-22, new-relic-probe-3) — this is the reverse direction (New Relic as MCP server, not MCP client). No evidence indicates New Relic's own AI assistant or platform can consume/plug in external MCP servers to use their tools.

   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”
   • [probe] “official MCP server documented at https://docs.newrelic.com/whats-new/2025/11/whats-new-11-05-mcp-server/”
   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”

   proof ↗⚑ flag

   Honeycomb—none0/10

   All MCP evidence describes Honeycomb acting as an MCP *server* that other AI agents connect to in order to query Honeycomb's own telemetry data (honeycomb-docs-9, honeycomb-docs-10, honeycomb-probe-3) — the opposite of the story, which asks whether Honeycomb itself can plug in external MCP servers to use their tools. No evidence shows Honeycomb consuming or hosting third-party MCP tools as a client.

   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”

   proof ↗⚑ flag

4. ai-native userConnect an agent via an official MCP server

   weight 3 · round drawn
   New Relic✓fullprobed8/10

   New Relic ships an official MCP server (documented Nov 2025) that connects AI development tools to New Relic's platform context, including converting plain-English questions into NRQL queries. This is first-party vendor documentation with a dedicated release note and probe-confirmed URL, but lacks independent/hands-on corroboration of the MCP server's reliability. Missing for 10: independent or hands-on validation of the MCP server, detail on setup/auth flow and supported tool integrations.

   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”
   • [probe] “official MCP server documented at https://docs.newrelic.com/whats-new/2025/11/whats-new-11-05-mcp-server/”

   proof ↗⚑ flag

   Honeycomb✓fullprobed8/10

   Honeycomb ships an official MCP server integration allowing any MCP-compatible AI agent to query live telemetry, investigate latency/error spikes, and translate dashboards/alerts into Honeycomb's query language, documented with a dedicated configuration guide and concepts page. Missing for 10: independent hands-on third-party validation of the MCP server itself (community evidence discusses agent value generally but not this specific MCP server in practice).

   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes; Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Translate existing dashboards and alerts into Honeycomb's query language”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”

   proof ↗⚑ flag

5. ai-native userUse an official CLI

   weight 2 · round to New Relic
   New Relic~partialprobed6/10

   New Relic ships an official open-source CLI (newrelic-cli) supporting entity search, deployment markers, and platform configuration, confirmed via GitHub repo and probe. However, the evidence doesn't show AI-native features specific to the CLI itself (e.g., natural-language query support, agentic workflows are only documented for the separate MCP server/AI assistant, not the CLI). Missing for 10: documentation of AI-native/agentic capabilities within the CLI itself, independent hands-on reviews of the CLI, and details on how it integrates with AI coding agents.

   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [probe] “official CLI documented at https://github.com/newrelic/newrelic-cli”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Evidence covers OpenTelemetry instrumentation, REST API, and an official MCP server, but no official CLI tool for Honeycomb is documented anywhere in the pack. Missing for 10: any mention of an official Honeycomb CLI, its installation, commands, or AI-native workflow usage.

   ⚑ flag

6. ai-native userDrive the product through a documented public API

   weight 3 · round drawn
   New Relic✓fullprobed8/10

   New Relic documents NerdGraph, a GraphQL API for querying data and configuring features, with an interactive explorer, plus an official CLI (newrelic-cli) that wraps API functionality for entity search and deployment markers. This constitutes a documented public API surface that can be driven programmatically outside the UI. Missing for 10: no discoverable OpenAPI/Swagger spec (probe found 404s) and no independent hands-on developer corroboration of NerdGraph usage beyond docs.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”
   • [probe] “official CLI documented at https://github.com/newrelic/newrelic-cli”

   proof ↗⚑ flag

   Honeycomb✓fullprobed8/10

   Honeycomb documents a public API for programmatically managing datasets, queries, triggers, SLOs, environments, and API keys, with a downloadable OpenAPI spec for tooling integration, which directly supports AI-native/agentic control of the product. Missing for 10: a live-hosted OpenAPI/swagger endpoint (probe found 404s on common paths) and independent third-party corroboration of API robustness beyond vendor docs.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

7. ai-native userIssue scoped/least-privilege API credentials for an agent

   weight 2 · round to Honeycomb
   New Relic—none0/10

   New Relic offers API access (NerdGraph, user keys) and an MCP server for AI tools, but no evidence describes scoped/least-privilege credential issuance specifically for agents (e.g., fine-grained API key scoping, role-based agent tokens). Missing for 10: documentation of scoped API key creation, least-privilege permission models for agent credentials, or agent-specific token management.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed4/10

   Honeycomb's API supports programmatic management of API keys and other resources (docs-11, docs-27), and agents can connect via the official MCP integration (docs-9), implying credential-based access, but there is no explicit documentation of scoped/least-privilege permission levels for API keys or agent-specific credential scoping. missing for 10: explicit docs on creating role-restricted or scoped API keys, least-privilege permission tiers, or agent-specific credential issuance workflow.

   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

8. ai-native userBuild against official SDKs

   weight 2 · round to New Relic
   New Relic✓fullprobed7/10

   New Relic ships official first-party agents/SDKs for Go, Java, .NET, Node.js, PHP, Python, and Ruby, plus a GraphQL-based NerdGraph API with an interactive API explorer, OpenTelemetry instrumentation support, and an official CLI (newrelic-cli), giving developers multiple official, documented ways to build against the platform. Missing for 10: independent/hands-on corroboration of SDK developer experience and no explicit AI-native SDK features (e.g., no OpenAPI spec found, llms.txt 404).

   • [claimed-docs] “Monitor application performance across your stack. We've got agents for Go, Java, .NET, Node.js, PHP, Python, and Ruby.”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “Instrument your service with OpenTelemetry and gain platform-agnostic observability.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [probe] “official CLI documented at https://github.com/newrelic/newrelic-cli”
   • [probe] “PROBE llms.txt: HTTP 404 at https://docs.newrelic.com/llms.txt”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”

   proof ↗⚑ flag

   Honeycomb~partialprobed6/10

   Honeycomb relies on OpenTelemetry SDKs (open standard, not Honeycomb-proprietary) for instrumentation, plus a REST/OpenAPI-based Honeycomb API for managing datasets, queries, triggers, and SLOs — this gives AI-native builders programmatic access but not a dedicated first-party 'Honeycomb SDK' in multiple languages. missing for 10: dedicated official Honeycomb-branded SDKs (vs generic OTel libraries), working OpenAPI spec download link (probe found 404s on common paths), and independent/hands-on developer corroboration of SDK build experience.

   • [claimed-docs] “Instrument your applications with OpenTelemetry, the open-source standard for collecting telemetry, and send traces, logs, and metrics to Ho…”
   • [claimed-docs] “Honeycomb supports receiving telemetry data via OpenTelemetry’s native protocol, OTLP, over gRPC, HTTP/protobuf, and HTTP/JSON.”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “This example creates a span, adds context as attributes, and closes it when the work is done”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

9. ai-native userSubscribe to events via webhooks

   weight 2 · round to Honeycomb
   New Relic—none0/10

   The evidence pack shows alerts, dashboards, NerdGraph API, MCP server, and CLI capabilities, but no mention of webhook subscriptions for events or any pub/sub-style event notification mechanism for external systems. Alerts docs mention notifications but do not describe webhook subscription APIs. Missing for 10: any documentation of a webhooks feature, webhook configuration API, or event subscription endpoint.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”

   proof ↗⚑ flag

   Honeycomb✓fullclaimed7/10

   Honeycomb lets users configure custom webhooks that receive JSON payloads whenever Triggers or SLO alerts fire, effectively enabling event subscription via webhooks alongside Slack/PagerDuty/Teams routing. Missing for 10: evidence of webhook subscriptions for a broader range of event types beyond triggers/SLOs, and independent/hands-on confirmation of webhook reliability for AI-native automation use cases.

   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”
   • [claimed-docs] “This allows you to build custom integrations that receive JSON payloads from Honeycomb upon alerts firing.”
   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”

   proof ↗⚑ flag

Agentic features

1. ai-native userGet AI-generated insights and suggestions from my data inside the product

   weight 2 · round to New Relic
   New Relic✓fullclaimed7/10

   New Relic AI (docs-9, docs-10) provides in-product AI assistant for plain-language querying, troubleshooting suggestions, and query help, plus an MCP server (docs-11, docs-22) that converts natural language to NRQL for insights. This is first-party documented but lacks independent/hands-on corroboration of quality or depth of insights generated. Missing for 10: independent user validation of AI insight quality, detail on proactive/automated insight generation beyond query assistance, and depth of anomaly-detection or automated root-cause AI features.

   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed6/10

   Honeycomb's Query Assistant translates natural-language questions into queries (AI-assisted analysis) and its MCP integration explicitly lets an AI agent 'investigate and diagnose latency or error spikes' and 'identify performance outliers and suggest optimization opportunities' using live Honeycomb data. However, the deeper insight/suggestion generation is delivered via an external MCP-connected agent rather than a fully native, always-on in-product AI insights panel, and BubbleUp (outlier detection) is mentioned only in pricing without AI framing. Missing for 10: evidence of a built-in AI-generated insights/summary feature independent of MCP agents, and independent/hands-on validation of suggestion quality.

   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes; Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “BubbleUp ![Checkmark]”

   proof ↗⚑ flag

2. ai-native userSet up automations that run autonomously in the background

   weight 2 · round drawn
   New Relic~partialclaimed4/10

   New Relic's alerting system lets you define thresholds and notification rules that run continuously in the background without further manual intervention (new-relic-docs-5, new-relic-docs-21/25), which is a basic autonomous background automation. However, the 'ai-native/agentic' framing (AI assistant, MCP server) is described as an interactive query/troubleshooting helper (new-relic-docs-9/10) or a connector for external AI tools (new-relic-docs-11/22) rather than an autonomous agent that itself sets up and runs background tasks. Missing for 10: evidence of AI-driven workflows/agents that autonomously configure or execute actions (not just user-defined alert rules), a documented automation/workflow engine beyond alerting, and independent corroboration that these run unattended reliably.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed4/10

   Honeycomb supports background automations via Triggers and SLOs that continuously evaluate conditions and fire alerts, plus an API to programmatically manage these automations — these run autonomously without user intervention. However, there is no evidence of AI-driven or agentic automation that acts autonomously (e.g., an agent scheduling investigations, auto-remediating, or running background tasks); the MCP integration is interactive (agent queries on request) rather than an autonomous background process. Missing for 10: evidence of AI/agent-initiated autonomous background workflows, scheduled agentic tasks, or autonomous remediation beyond static threshold-based triggers/SLOs.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”

   proof ↗⚑ flag

3. ai-native userDelegate tasks to a built-in AI assistant inside the product

   weight 3 · round to New Relic
   New Relic✓fullclaimed7/10

   New Relic docs describe a built-in AI assistant (New Relic AI) that lets users ask questions in plain language, troubleshoot issues, explore telemetry data, and get help building NRQL queries directly inside the product. This matches the story of delegating tasks to an in-product AI assistant, though evidence is vendor-documentation only with no independent hands-on verification of task delegation depth or reliability. Missing for 10: independent/community corroboration of the assistant's real-world task delegation capability and quality, and more detail on the scope of tasks it can autonomously perform beyond Q&A/query help.

   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed4/10

   Honeycomb ships a built-in 'Query Assistant' that lets users generate queries from natural language input, a narrow built-in AI feature — but this is limited to query construction, not general task delegation (investigation, remediation, cross-tool reasoning), which docs instead offload to an external MCP-connected agent. Missing for 10: evidence of a built-in assistant that can autonomously investigate issues, take multi-step actions, or operate beyond query generation within the product itself.

   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”

   proof ↗⚑ flag

4. ai-native userOperate the product with natural-language commands

   weight 2 · round drawn
   New Relic✓fullprobed8/10

   New Relic AI lets users ask questions, troubleshoot, and explore telemetry in plain language, and the new MCP server explicitly converts natural-language questions into NRQL queries to retrieve results without requiring NRQL expertise. This directly delivers natural-language operation of the product for querying and troubleshooting. Missing for 10: independent/hands-on validation of the natural-language assistant's accuracy and broader coverage beyond querying (e.g., configuring alerts or dashboards via NL commands).

   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [probe] “official MCP server documented at https://docs.newrelic.com/whats-new/2025/11/whats-new-11-05-mcp-server/”

   proof ↗⚑ flag

   Honeycomb✓fullprobed8/10

   Honeycomb's Query Assistant lets users generate queries from natural-language input directly in the product, and the official MCP server lets AI agents query telemetry, diagnose issues, and translate dashboards/alerts using natural-language interaction with live data (honeycomb-docs-4, honeycomb-docs-9, honeycomb-docs-10, honeycomb-docs-23, honeycomb-probe-3). Missing for 10: independent/hands-on evidence validating NLQ accuracy and broader coverage of natural-language commands beyond querying (e.g., configuring triggers/SLOs via NL).

   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Translate existing dashboards and alerts into Honeycomb's query language”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”

   proof ↗⚑ flag

Api quality

1. ai-native userExplore an interactive API reference with runnable examples

   weight 2 · round to New Relic
   New Relic~partialprobed5/10

   New Relic documents a NerdGraph API explorer for experimenting with its GraphQL API (new-relic-docs-20), which functions as an interactive, runnable reference for that API, but there's no evidence of a broader interactive API reference (e.g., REST API explorer/Swagger UI) covering the full platform, and probes show no OpenAPI spec or llms.txt exposed. Missing for 10: a comprehensive interactive reference across all New Relic APIs (not just NerdGraph), OpenAPI/Swagger-based explorer, and independent/hands-on confirmation of the explorer's usability.

   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Docs mention an API and a downloadable OpenAPI spec, but there is no evidence of an interactive, browsable API reference with runnable/try-it examples; a probe for common OpenAPI/swagger endpoints returned 404s, further indicating no discoverable interactive reference.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

2. ai-native userDownload a machine-readable API spec (OpenAPI or equivalent)

   weight 2 · round to Honeycomb
   New Relic—none0/10

   New Relic's primary API is NerdGraph (GraphQL), and explicit probes show no OpenAPI/swagger spec is served at any standard path (404s), with no other evidence of a downloadable machine-readable spec.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”

   proof ↗⚑ flag

   Honeycomb✓fullprobed8/10

   Honeycomb docs explicitly state you can download the Honeycomb OpenAPI spec for use with your own tooling, directly satisfying the story, though a live probe of common OpenAPI spec URLs returned 404s rather than confirming an easily discoverable public endpoint. Missing for 10: independent/hands-on confirmation that the spec is actually reachable at a stable public URL, and details on spec completeness/versioning.

   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

3. ai-native userTest against a sandbox environment without touching production data

   weight 1 · round drawn
   New Relic—none0/10

   No evidence in the pack of a sandbox/test environment, free trial account isolation, or any mechanism to test New Relic's features without touching production telemetry data; all references are to production monitoring, alerting, dashboards, and the MCP server connecting to real platform data. Missing for 10: any documented sandbox account, test-mode API, synthetic data environment, or explicit separation between test and production data ingestion.

   ⚑ flag

   Honeycomb—none0/10

   Honeycomb is an observability platform with a free tier and environments, but the evidence pack contains no mention of a sandbox/test environment for AI agents to safely experiment against without touching production telemetry data. Environments are mentioned only in passing (API key management), with no documented sandbox mode or synthetic-data test environment.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

4. ai-native userRely on versioned APIs with a documented deprecation policy

   weight 2 · round drawn
   New Relic—none0/10

   Evidence shows NerdGraph API and NerdGraph explorer exist, but there is no documented API versioning scheme or deprecation policy; probes confirm no OpenAPI spec is discoverable, and no docs mention version numbers, sunset timelines, or breaking-change guarantees. Missing for 10: documented API versioning scheme, explicit deprecation/sunset policy, changelog or migration guide for breaking changes.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Evidence shows an API and OpenAPI spec exist (honeycomb-docs-21, honeycomb-docs-27) but there is no mention of API versioning scheme or a documented deprecation policy anywhere in the pack, and a probe even failed to find an openapi.json at expected locations.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

Agent integration

1. ai-native userHave an external agent query metrics, logs, and traces through documented APIs to debug production

   weight 3 · round to New Relic
   New Relic✓fullprobed9/10

   New Relic exposes NerdGraph, a documented GraphQL API for querying metrics, logs, and traces, plus NRQL for structured queries, and now an official MCP server that lets external AI agents connect and convert natural-language questions into NRQL for troubleshooting production issues. This directly matches the story of an external agent querying telemetry through documented APIs to debug production. Missing for 10: independent hands-on validation of the MCP server/API reliability from third parties beyond vendor docs.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”
   • [probe] “official MCP server documented at https://docs.newrelic.com/whats-new/2025/11/whats-new-11-05-mcp-server/”

   proof ↗⚑ flag

   Honeycomb✓fullprobed8/10

   Honeycomb documents an official MCP server that lets any external AI agent query traces/metrics and investigate issues using live Honeycomb data, plus a full REST API (with OpenAPI spec) for programmatic query/dataset/trigger management, and community reports confirm agents cross-referencing Honeycomb traces with other telemetry sources during incidents. missing for 10: independent hands-on verification of the MCP server's reliability/quality, and a directly reachable OpenAPI spec URL (probe found 404s on standard paths).

   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [community] “One of the interesting things an agent can do that no individual telemetry tool does effectively is make deductions and integrate informatio…”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

Ai investigation

1. ai-native userGet AI-generated summaries of incidents and alert context for responders

   weight 2 · round drawn
   New Relic~partialclaimed5/10

   New Relic AI (docs-9, docs-10) lets users ask questions and troubleshoot issues in plain language, and the MCP server (docs-11, docs-22) exposes platform context to AI tools, which supports AI-assisted incident triage — but no evidence explicitly shows auto-generated incident summaries or alert-specific responder context. Missing for 10: explicit incident-summary generation feature, alert-context packaging for responders, and independent/hands-on confirmation of this specific workflow.

   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”

   proof ↗⚑ flag

   Honeycomb~partialprobed5/10

   Honeycomb's MCP server lets connected AI agents investigate/diagnose latency or error spikes and suggest optimizations using live telemetry, which can produce incident/alert context summaries, and Query Assistant translates natural language into queries — but this is agent-mediated rather than a native built-in 'incident summary' feature for responders. Missing for 10: a first-party, no-agent-required feature that automatically generates written incident summaries or alert-context narratives, and any hands-on/community evidence validating summary quality for responders.

   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes; Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Translate existing dashboards and alerts into Honeycomb's query language”
   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”

   proof ↗⚑ flag

2. ai-native userHave the platform's AI investigate an alert or error and propose a probable root cause

   weight 3 · round to Honeycomb
   New Relic~partialclaimed5/10

   New Relic AI is documented to let users 'ask questions, troubleshoot issues, and explore telemetry data using plain language' and the MCP server converts natural-language questions into NRQL queries, which supports AI-assisted investigation of issues, but the evidence never explicitly describes an AI that proactively investigates a specific alert/error and outputs a proposed root cause. Missing for 10: explicit documentation of an automated root-cause-analysis workflow tied to alerts/errors, hands-on/independent confirmation that the AI assistant produces accurate root-cause conclusions rather than just query help.

   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”

   proof ↗⚑ flag

   Honeycomb~partialprobed6/10

   Honeycomb's MCP integration lets connected AI agents query live telemetry, investigate/diagnose latency or error spikes, and identify performance outliers with optimization suggestions—close to proposing a probable root cause—but this requires an external MCP-compatible AI agent rather than a fully native, built-in 'Honeycomb AI' feature, and the docs stop short of explicit 'root cause' language. Missing for 10: a native (non-MCP-dependent) AI root-cause proposal feature, explicit 'root cause' framing, and independent/hands-on evidence of investigation accuracy.

   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes; Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Translate existing dashboards and alerts into Honeycomb's query language”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”
   • [community] “One of the interesting things an agent can do that no individual telemetry tool does effectively is make deductions and integrate informatio…”

   proof ↗⚑ flag

Ai querying

1. ai-native userAsk questions of my telemetry in natural language and get a real query or chart back

   weight 2 · round drawn
   New Relic✓fullclaimed8/10

   New Relic AI explicitly lets users ask questions, troubleshoot, and explore telemetry in plain language, and can help generate/fix NRQL queries; the new MCP server converts natural-language questions directly into NRQL and returns results, tying into chart/dashboard capabilities. missing for 10: independent/hands-on validation of natural-language-to-chart accuracy and no community corroboration of the AI assistant's real-world reliability.

   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”
   • [claimed-docs] “Ask questions in plain English, and the MCP server converts them to NRQL to get results—even if you're not a NRQL expert.”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”
   • [claimed-docs] “Through our UI you can build and share tailored user-friendly charts and visualizations.”

   proof ↗⚑ flag

   Honeycomb✓fullprobed8/10

   Honeycomb's Query Assistant explicitly generates Honeycomb queries from natural-language input, directly matching the story, and the MCP integration extends this so AI agents can query telemetry and get real answers/charts back. missing for 10: independent/hands-on validation of Query Assistant's accuracy and no evidence of chart-specific output beyond query generation.

   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [claimed-docs] “Connect Honeycomb to any AI agent that supports the Model Context Protocol (MCP) so it can query your telemetry, investigate issues, and ans…”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Translate existing dashboards and alerts into Honeycomb's query language”
   • [probe] “official MCP server documented at https://docs.honeycomb.io/integrations/mcp/concepts”

   proof ↗⚑ flag

Alert automation

1. ai-native userPoint alert notifications at webhooks that trigger automated remediation or agents

   weight 2 · round to Honeycomb
   New Relic—none0/10

   The evidence pack describes generic alerting configuration (thresholds, notification recipients) and a GraphQL API (NerdGraph) for configuring some features, but nowhere explicitly mentions webhook notification channels or integration with remediation/automation agents via webhooks. Given New Relic is an observability/alerting platform, this is a fair and applicable axis, but no direct evidence confirms webhook-triggered automation exists.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”

   proof ↗⚑ flag

   Honeycomb✓fullclaimed7/10

   Honeycomb's Triggers and SLOs can route alert notifications to custom webhooks that receive JSON payloads on firing, which is exactly the mechanism needed to trigger automated remediation scripts or agents [honeycomb-docs-8][honeycomb-docs-18]. The API also allows programmatic management of triggers/SLOs for building such integrations [honeycomb-docs-11][honeycomb-docs-27]. Missing for 10: a concrete documented example of a webhook wired to an automated remediation workflow or AI agent, and independent/community confirmation that this webhook-to-agent pattern works in practice.

   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”
   • [claimed-docs] “This allows you to build custom integrations that receive JSON payloads from Honeycomb upon alerts firing.”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

Alerting

1. sreAlert on any telemetry signal with routing, grouping, and silencing of notifications

   weight 3 · round to Honeycomb
   New Relic~partialclaimed5/10

   New Relic docs confirm NRQL-based alerting on any telemetry signal and thresholds with notification routing ("who is notified and how"), but the evidence pack contains no explicit mention of alert grouping or silencing/muting mechanisms. Missing for 10: documented grouping of related alerts, muting/silencing rules, and incident routing workflows with independent corroboration.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”

   proof ↗⚑ flag

   Honeycomb~partialclaimed6/10

   Honeycomb documents Triggers and SLO-based alerts that fire on threshold/burn-rate conditions and route notifications to Slack, PagerDuty, Microsoft Teams, or custom webhooks, covering alerting on telemetry signals and routing. However, the evidence pack contains no explicit documentation of alert grouping (deduplication/aggregation) or silencing/muting of notifications. missing for 10: explicit docs on notification grouping/deduplication, silencing or snoozing alerts.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”
   • [claimed-docs] “This allows you to build custom integrations that receive JSON payloads from Honeycomb upon alerts firing.”

   proof ↗⚑ flag

2. sreEnable anomaly or outlier detection that surfaces problems without hand-written thresholds

   weight 1 · round to Honeycomb
   New Relic—none0/10

   The evidence pack describes only threshold-based alerting ('you define the thresholds that if exceeded mean an issue') and NRQL-based alerts, with no mention of anomaly/outlier/baseline detection features that surface issues without hand-written thresholds.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”

   proof ↗⚑ flag

   Honeycomb~partialclaimed4/10

   Honeycomb's core alerting mechanism (Triggers, SLOs) is explicitly threshold-based (docs-6,7,8), not automated anomaly detection. BubbleUp is mentioned only in a pricing checklist (docs-28) with no doc detail on how it works, and the MCP-connected AI agent can 'identify performance outliers' (docs-10/20/26) but this is an on-demand investigative query tool, not a standing anomaly-detection alert that surfaces problems without human-defined thresholds. Missing for 10: documented automatic/passive anomaly-detection alerting, technical detail on BubbleUp's outlier algorithm, and evidence it triggers notifications without manual threshold configuration.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “BubbleUp ![Checkmark]”
   • [claimed-docs] “They can: * Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes; Identify performance outliers and suggest optimization opportunities”
   • [claimed-docs] “Investigate and diagnose latency or error spikes * Identify performance outliers and suggest optimization opportunities”

   proof ↗⚑ flag

Slos

1. sreDefine SLOs with error budgets and burn-rate alerts

   weight 2 · round drawn
   New Relic✓fullclaimed8/10

   New Relic's Service Level Management explicitly supports defining and consuming SLIs/SLOs, and NRQL-based alerts (its 'primary and most powerful' alert type) can be layered to implement burn-rate style alerting on top of SLO data. Missing for 10: explicit first-party documentation of a native 'error budget burn-rate alert' template/wizard (as some competitors have) and independent/hands-on confirmation of this specific workflow beyond docs.

   • [claimed-docs] “With New Relic, you can define and consume service level indicators (SLIs) and service level objectives (SLOs) for your applications.”
   • [claimed-docs] “you can define and consume service level indicators (SLIs) and service level objectives (SLOs) for your applications”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”

   proof ↗⚑ flag

   Honeycomb✓fullclaimed8/10

   Honeycomb docs explicitly document SLOs with error budgets ('be alerted when your SLO budget is threatened'), Budget Burndown/Historical Compliance tracking, and burn-rate style alerting routed to Slack/PagerDuty/Teams/webhooks, plus API support for programmatic SLO management. Missing for 10: no independent/hands-on validation of alert accuracy or burn-rate tuning specifics beyond docs.

   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”
   • [claimed-docs] “Honeycomb will track SLO values past your retention period, but will display these for only the Budget Burndown and Historical Compliance gr…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

1. ai-native userPerform bulk operations across many items at once

   weight 2 · round to New Relic
   New Relic~partialclaimed5/10

   NerdGraph GraphQL API and the newrelic-cli support programmatic/bulk operations like entity search across all accounts, and NRQL/dashboards can operate over aggregated data sets, but there's no explicit documentation of bulk create/update/delete operations across many items (e.g., bulk alert config, bulk entity tagging) as a first-class feature. missing for 10: explicit bulk-operation endpoints/examples (batch create/update/delete across many entities), hands-on evidence of bulk workflows succeeding at scale.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb's API docs mention programmatic management of datasets, queries, triggers, SLOs, etc. (honeycomb-docs-11, honeycomb-docs-27), but no evidence describes batch/bulk endpoints or bulk-edit/delete workflows across many items at once; the docs only reference singular resource management.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

2. ai-native userDefine rules that trigger actions automatically on events

   weight 3 · round drawn
   New Relic✓fullclaimed7/10

   New Relic's alerting system lets users define thresholds/conditions (including powerful NRQL-based conditions) and specify who/how to notify when those conditions are breached, which is the core of automated rule-triggered actions on events (new-relic-docs-5, new-relic-docs-3/21/25). missing for 10: evidence of richer automated remediation/workflow actions (e.g., auto-scaling, ticket creation, custom webhooks) beyond notification routing, and independent/hands-on confirmation that alert rules reliably trigger actions in production.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”

   proof ↗⚑ flag

   Honeycomb✓fullclaimed7/10

   Honeycomb's Triggers let users define threshold-based rules that automatically fire alerts/actions (Slack, PagerDuty, Teams, or custom webhook) when conditions are met, and SLOs similarly alert on budget breaches — directly matching the 'rules trigger actions on events' pattern. The API also allows programmatic management of triggers for automated workflows. Missing for 10: AI-assisted or natural-language rule authoring specifically, more complex conditional/chained automation logic, and independent/hands-on validation of trigger reliability beyond vendor docs.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”
   • [claimed-docs] “This allows you to build custom integrations that receive JSON payloads from Honeycomb upon alerts firing.”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

3. ai-native userSchedule recurring jobs or workflows

   weight 2 · round drawn
   New Relic—none0/10

   New Relic's evidence pack shows synthetic monitoring (docs-19) and alerting/dashboards, but there is no documentation of a general job/workflow scheduler, cron-like automation, or AI-native recurring workflow execution capability tied to its MCP server or NerdGraph API. The axis is plausible for a monitoring platform (e.g., scheduled synthetic checks), but no evidence confirms recurring job/workflow scheduling as a user-facing automation feature.

   • [claimed-docs] “Simulate end-user activity and API calls from locations around the world, so you can find problems before your users do.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb documents triggers/SLO alerts and an API for automation, but nothing about scheduling recurring jobs, reports, or workflows to run on a cadence — the evidence covers threshold-based alerting and ad-hoc API automation, not scheduled/recurring execution.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”

   proof ↗⚑ flag

4. ai-native userVersion, review, and roll back my automations

   weight 1 · round drawn
   New Relic—none0/10

   The evidence pack shows NerdGraph API and CLI for configuring alerts/dashboards, but nothing about versioning, reviewing, or rolling back automations (e.g., alert configs, workflows) — no changelog, diff/review UI, or rollback mechanism is documented.

   ⚑ flag

   Honeycomb—none0/10

   Honeycomb offers Boards, Triggers, SLOs, and an API to manage configuration, but there is no evidence of version control, review workflows, or rollback capability for automations (triggers/SLOs/boards) — no changelog, diff, or revert feature is documented.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”

   proof ↗⚑ flag

Cost

1. sreSee what my observability spend is, attribute it to teams or services, and catch usage spikes before the bill

   weight 3 · round drawn
   New Relic—none0/10

   There is no evidence of a New Relic feature for attributing observability ingest/spend to specific teams or services, nor of proactive spend/usage-spike alerting on New Relic's own billing (the closest hit, new-relic-docs-15, is about monitoring customers' cloud/K8s spend, not New Relic's own consumption). Multiple community reports directly describe the opposite experience — unexpected multi-x ingest spikes and bill surges that were only discovered after the fact (new-relic-comm-4, new-relic-comm-5, new-relic-comm-8, new-relic-comm-9), with no mention of a dashboard or alert that would have caught them beforehand.

   • [claimed-docs] “Gain real-time visibility into multi-cloud and Kubernetes spend.”
   • [community] “New Relic wants to capture as much data as possible so that they can charge you as much as possible. I run a small environment (~20 small VM…”
   • [community] “Reminds me of when they intentionally enabled JFR on every single java process. Roughly sextupled our ingest (and associated costs) until we…”
   • [community] “I love NewRelic, but their pricing always baffled me. When we added it to our QA, UAT, and Staging envs, they started wanting to bill us as …”
   • [community] “We also run our infrastructure with redundancy, so we got billed for every region & availability zone, they wanted it to be crazy expensive.”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Evidence shows only flat pricing tiers with volume caps (event/metrics limits) but nothing about per-team/service cost attribution, spend dashboards, or usage-spike alerting tied to billing; Triggers/SLOs in the pack are about reliability, not cost governance.

   • [claimed-docs] “Our introductory plan, free forever.”
   • [claimed-docs] “Event Volume: Up to 20M per month Metrics Data Points: Up to 100M per month”
   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”

   proof ↗⚑ flag

2. srePredict costs from transparent published per-signal pricing without talking to sales

   weight 1 · round to Honeycomb
   New Relic!disputedcontradicted4/10

   New Relic does publish some pricing info (100GB free ingest tier) on its pricing page, giving a baseline for self-service cost estimation without sales contact, but multiple independent community reports describe billing as opaque and unpredictable in practice—per-user pricing changes causing 20x bill increases, environments/redundant AZs billed as extra 'production' instances, and ingest spikes (e.g. JFR enabled by default) sextupling costs unexpectedly. This directly contradicts the 'predict costs transparently' premise even though a published price list exists. Missing for 10: detailed per-signal (e.g., per-host, per-user, per-GB-by-data-type) pricing breakdown, and evidence that published prices reliably predict actual bills without support intervention.

   • [claimed-docs] “100 GB of free data ingest/month”
   • [claimed-docs] “Sign up for free access with our perpetual free tier and 100 GB/month of data ingest included.”
   • [community] “When they changed their billing to be per user our bill went up 20x for no additional value... APM hasn't seen a single new feature in the 5…”
   • [community] “New Relic wants to capture as much data as possible so that they can charge you as much as possible. I run a small environment (~20 small VM…”
   • [community] “Reminds me of when they intentionally enabled JFR on every single java process. Roughly sextupled our ingest (and associated costs) until we…”
   • [community] “I love NewRelic, but their pricing always baffled me. When we added it to our QA, UAT, and Staging envs, they started wanting to bill us as …”
   • [community] “We also run our infrastructure with redundancy, so we got billed for every region & availability zone, they wanted it to be crazy expensive.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed5/10

   Honeycomb's public pricing page is referenced with a free tier and explicit volume caps (20M events/mo, 100M metric data points/mo) suggesting some self-serve, published tiers, but the evidence never shows actual per-signal dollar rates or a cost calculator, and higher tiers likely require sales contact. Missing for 10: explicit $/event or $/metric pricing figures, confirmation that all tiers (including enterprise) are self-serve without sales engagement, and independent corroboration that published pricing lets an SRE fully predict costs.

   • [claimed-docs] “Our introductory plan, free forever.”
   • [claimed-docs] “BubbleUp ![Checkmark]”
   • [claimed-docs] “Event Volume: Up to 20M per month Metrics Data Points: Up to 100M per month”

   proof ↗⚑ flag

Sampling

1. developerControl trace/log sampling and retention tiers to manage data volume deliberately

   weight 2 · round drawn
   New Relic~partialcommunity4/10

   There is only indirect evidence: community reports mention disabling automatic log forwarding via env vars and manually turning off JFR-driven ingest spikes, implying some data-volume control exists, and the free tier caps ingest at 100GB/month. However, no first-party documentation describes explicit trace sampling rules, log drop filters, or configurable data retention tiers as a deliberate developer-facing feature. missing for 10: dedicated docs on trace/log sampling rate configuration, NRQL drop-filter rules, and retention tier selection; independent confirmation that these controls work as intended.

   • [community] “Setting NEW_RELIC_APPLICATION_LOGGING_ENABLED=false and NEW_RELIC_APPLICATION_LOGGING_FORWARDING_ENABLED=false disables the newly-default au…”
   • [community] “New Relic wants to capture as much data as possible so that they can charge you as much as possible. I run a small environment (~20 small VM…”
   • [community] “Reminds me of when they intentionally enabled JFR on every single java process. Roughly sextupled our ingest (and associated costs) until we…”
   • [claimed-docs] “100 GB of free data ingest/month”
   • [claimed-docs] “Sign up for free access with our perpetual free tier and 100 GB/month of data ingest included.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed4/10

   The pack shows plan-based volume tiers (event volume up to 20M/month) and references a data retention period that affects SLO history, implying retention/volume controls exist, but there is no documentation of actual sampling configuration (e.g., head/tail sampling, deterministic sampling rules) that a developer could set to deliberately manage trace/log volume. missing for 10: explicit sampling configuration docs (head/tail sampling, sample rate settings), explicit retention-tier management/configuration documentation, independent confirmation of sampling behavior in production.

   • [claimed-docs] “Honeycomb will track SLO values past your retention period, but will display these for only the Budget Burndown and Historical Compliance gr…”
   • [claimed-docs] “Our introductory plan, free forever.”
   • [claimed-docs] “Event Volume: Up to 20M per month Metrics Data Points: Up to 100M per month”

   proof ↗⚑ flag

As code

1. developerDefine dashboards and alerts as code (JSON models, Terraform, or API) and provision them repeatably

   weight 3 · round drawn
   New Relic~partialclaimed5/10

   NerdGraph is documented as a GraphQL API that can 'configure some New Relic features' and has an API explorer, implying dashboards/alerts could be provisioned programmatically, and the newrelic-cli exists for some automation tasks. However, the evidence pack never explicitly documents JSON dashboard definition files, a Terraform provider, or alerts-as-code workflows — dashboards are described mainly as built through the UI. Missing for 10: explicit Terraform provider docs, JSON dashboard schema/export-import examples, and alert-as-code API examples beyond generic NerdGraph mention.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “Through our UI you can build and share tailored user-friendly charts and visualizations.”
   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”

   proof ↗⚑ flag

   Honeycomb~partialclaimed5/10

   Honeycomb's API explicitly allows programmatic management of queries, triggers, SLOs, environments, and API keys, and offers a downloadable OpenAPI spec for custom tooling, supporting alerts-as-code provisioning [honeycomb-docs-11][honeycomb-docs-27][honeycomb-docs-21]. However, there is no evidence of a JSON dashboard schema or Terraform provider for Boards, nor confirmation that Boards (dashboards) themselves are manageable via the API—only 'Board Templates' are mentioned as a manual time-saver [honeycomb-docs-19][honeycomb-docs-5]. missing for 10: explicit Terraform provider/module, documented dashboard (Board) JSON schema or API endpoints for creating/updating Boards, and independent confirmation of repeatable provisioning workflows.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “we offer pre-configured Board Templates for common use cases... saving you time that would otherwise be spent building Boards manually.”
   • [claimed-docs] “A Board is a workspace where you can save, organize, and share analysis components focused on a common objective.”
   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”

   proof ↗⚑ flag

Dashboards

1. sreBuild shareable dashboards with rich visualization types and template variables

   weight 2 · round to New Relic
   New Relic~partialclaimed7/10

   Docs confirm New Relic dashboards support rich visualizations built via UI/NRQL and can be publicly shared via live dashboard URLs, satisfying most of the story. However, evidence does not explicitly mention template variables on dashboards, and there is no independent/hands-on corroboration of the dashboard-building experience (community evidence focuses on pricing/UI complaints unrelated to dashboard features). Missing for 10: explicit documentation of template variables support, independent verification of dashboard usability/quality.

   • [claimed-docs] “Through our UI you can build and share tailored user-friendly charts and visualizations.”
   • [claimed-docs] “Share dashboards with stakeholders outside your New Relic organization via publicly accessible live dashboard URLs.”
   • [claimed-docs] “Through our UI you can build and share tailored user-friendly charts and visualizations. Go to our dashboard capability to gather and chart …”
   • [claimed-docs] “Share dashboards with stakeholders outside your New Relic organization via publicly accessible live dashboard URLs”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”

   proof ↗⚑ flag

   Honeycomb~partialclaimed5/10

   Honeycomb Boards let SREs save, organize, and share analysis components, and pre-built Board Templates exist for common use cases, satisfying the shareable-dashboard and templating-for-speed aspect. However, evidence does not document dashboard 'template variables' (parameterized/dynamic dashboards) or a broad set of rich visualization types beyond query results and BubbleUp outlier views. Missing for 10: explicit template-variable support, documented visualization type gallery (e.g., heatmaps, gauges beyond BubbleUp), independent user validation of dashboard richness.

   • [claimed-docs] “A Board is a workspace where you can save, organize, and share analysis components focused on a common objective.”
   • [claimed-docs] “we offer pre-configured Board Templates for common use cases... saving you time that would otherwise be spent building Boards manually.”
   • [claimed-docs] “BubbleUp ![Checkmark]”

   proof ↗⚑ flag

Local dev

1. developerSpin up a local or dev instance of the platform to test instrumentation and dashboards

   weight 1 · round drawn
   New Relic—none0/10

   New Relic is delivered as a hosted SaaS platform; none of the docs describe a way to run a local or self-hosted dev instance, and community evidence explicitly notes the 'removal of development mode & support' as a past feature that is no longer offered, confirming there is no current local/dev instance capability.

   • [community] “Removal of development mode & support. The 'improved' price model. Their new UI is buggy as hell and they refuse to fix it. Every month or s…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb is described throughout as a cloud SaaS platform (OTLP ingestion, hosted query builder, boards, SLOs, API); no evidence of a local/self-hosted or on-prem dev instance for testing instrumentation and dashboards. The free tier (docs-13, docs-29) is still a hosted cloud account, not a local/dev deployment.

   ⚑ flag

Self host

1. sreRun the full observability stack self-hosted in production with documented architecture and upgrade path

   weight 2 · round drawn
   New Relic—none0/10

   New Relic is documented as a SaaS platform (agents forward data to New Relic-hosted cloud) with no self-hosted deployment architecture, upgrade path, or on-prem installation docs anywhere in the evidence. Community commentary explicitly notes New Relic controls 'where the server is' and that on-prem APM remains an underserved market they don't address, confirming no self-hosted offering exists.

   • [community] “They don’t care that much about open source agent, but they do care about where the server is... there is still a huge underserved market fo…”
   • [claimed-docs] “By default, our APM agents do two things: add metadata to your logs, which gives you logs in context (ability to see logs data in various re…”
   • [claimed-docs] “100 GB of free data ingest/month”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb is a SaaS observability product; no evidence pack item mentions self-hosting, on-prem deployment, or an upgrade path for a self-managed stack — all docs reference Honeycomb's own hosted platform and pricing tiers. This is an applicable axis (self-hosted observability stacks exist as a category, e.g. open-source alternatives) but Honeycomb offers no documented self-hosted deployment option.

   • [claimed-docs] “Our introductory plan, free forever.”
   • [claimed-docs] “Event Volume: Up to 20M per month Metrics Data Points: Up to 100M per month”

   proof ↗⚑ flag

Change tracking

1. developerCorrelate regressions with deploys and configuration changes via release or change tracking

   weight 2 · round to New Relic
   New Relic~partialclaimed6/10

   New Relic's CLI supports recording deployment markers (change tracking) that appear alongside APM data, and NRQL/dashboards/alerts can be used to correlate metrics around deploy times, but the evidence pack lacks documentation of a dedicated change-tracking UI feature, automatic regression detection tied to deploys, or config-change tracking beyond deployment markers. missing for 10: dedicated change-tracking/regression-correlation UI docs, evidence of automatic config-change tracking, independent confirmation of deploy correlation workflow.

   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   The evidence pack covers OpenTelemetry ingestion, querying, boards, triggers/SLOs, and MCP integration, but nowhere mentions Honeycomb's deploy/release marker or annotation feature that would let a developer explicitly correlate regressions with deploys or config changes; this is a standard, fair capability for an observability platform, so its absence is a gap rather than an inapplicable axis.

   ⚑ flag

Incidents

1. sreDeclare and track incidents with timelines, on-call schedules, and escalation policies

   weight 2 · round to New Relic
   New Relic~partialclaimed3/10

   New Relic's alerting docs mention defining thresholds and who is notified (new-relic-docs-5, new-relic-docs-25), which touches escalation/notification, but there is no evidence of a dedicated incident timeline view, on-call scheduling, or formal escalation-policy management comparable to a dedicated incident-response tool. Missing for 10: documented incident timeline tracking, on-call schedule configuration, and escalation policy chains/rotations.

   • [claimed-docs] “You define the data to watch, the thresholds that if exceeded mean an issue, who is notified, and how.”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb provides triggers and SLO alerts that route notifications to PagerDuty/Slack/Teams, but there is no evidence Honeycomb itself declares incidents, tracks incident timelines, manages on-call schedules, or defines escalation policies — those are delegated to external tools like PagerDuty.

   • [claimed-docs] “Use triggers to send alerts when thresholds that you define and configure are passed.”
   • [claimed-docs] “Use Service Level Objectives (SLOs) to define an agreement regarding delivery of a given service and be alerted when your SLO budget is thre…”
   • [claimed-docs] “Set up Triggers and SLOs to alert your team when conditions are met, and route notifications to Slack, PagerDuty, Microsoft Teams, or a cust…”

   proof ↗⚑ flag

1. ai-native userDo everything through the API that I can do in the UI

   weight 2 · round to Honeycomb
   New Relic~partialprobed5/10

   NerdGraph GraphQL API and CLI expose query, alerting, dashboard, and SLO configuration, but New Relic's own docs state NerdGraph lets you 'configure some New Relic features' — explicitly not full UI parity. No evidence every UI workflow (e.g., full dashboard builder, entity management edge cases) is API-exposed, and openapi probes 404 confirming no fully documented spec of complete surface. Missing for 10: explicit confirmation of 1:1 API/UI feature parity, comprehensive OpenAPI spec, and independent verification of coverage gaps.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.newrelic.com/openapi.json, https://docs.newrelic.com/swagger.json, https://docs.newreli…”
   • [probe] “official CLI documented at https://github.com/newrelic/newrelic-cli”

   proof ↗⚑ flag

   Honeycomb~partialprobed6/10

   Honeycomb's API lets users programmatically manage datasets, queries, triggers, SLOs, environments, and API keys (honeycomb-docs-11, honeycomb-docs-27), and an OpenAPI spec is downloadable (honeycomb-docs-21), covering much of the UI's core functionality. However, UI-only features like Query Assistant (NLQ), Boards/Board Templates, and BubbleUp are not documented as API-accessible, and a probe found no hosted OpenAPI spec at expected endpoints (honeycomb-probe-2), suggesting API parity may be incomplete or harder to discover. missing for 10: API-level access to Query Assistant/NLQ, Boards/Board Templates management via API, BubbleUp analysis via API, and a directly accessible OpenAPI spec confirming full endpoint coverage.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [claimed-docs] “A Board is a workspace where you can save, organize, and share analysis components focused on a common objective.”
   • [claimed-docs] “we offer pre-configured Board Templates for common use cases... saving you time that would otherwise be spent building Boards manually.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

2. ai-native userExport all of my data in open formats and leave

   weight 3 · round to New Relic
   New Relic~partialcommunity3/10

   New Relic exposes a GraphQL-based NerdGraph API and NRQL query language that could be used to pull data out programmatically, and it supports OpenTelemetry for ingestion, but there is no documented bulk 'export all my data' feature, no data-portability/exit tooling, and community comments note New Relic 'cares about where the server is' implying lock-in to their platform. missing for 10: dedicated bulk export/backup feature, explicit open-format export guarantee, independent confirmation of full data portability on exit.

   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “Instrument your service with OpenTelemetry and gain platform-agnostic observability.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [community] “They don’t care that much about open source agent, but they do care about where the server is... there is still a huge underserved market fo…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Evidence shows Honeycomb ingests data via the open OTLP/OpenTelemetry standard and offers an API/OpenAPI spec for managing datasets, queries, and configs, but nothing in the pack documents a bulk data-export mechanism for retrieving stored traces/logs/metrics in an open format to migrate away — the probe even shows no discoverable OpenAPI endpoint. Missing for 10: explicit bulk export/download capability for raw telemetry data, documented data-portability guarantees, and any evidence of exporting historical events rather than just querying or sending data in.

   • [claimed-docs] “Build integrations and automate workflows with the Honeycomb API. Programmatically manage datasets, queries, triggers, SLOs, environments, A…”
   • [claimed-docs] “You can download the Honeycomb OpenAPI spec to use with your own tooling.”
   • [claimed-docs] “Programmatically manage datasets, queries, triggers, SLOs, environments, API keys, and more.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://docs.honeycomb.io/openapi.json, https://docs.honeycomb.io/swagger.json, https://docs.honeyco…”

   proof ↗⚑ flag

3. ai-native userRead the product's source under an open license

   weight 2 · round drawn
   New Relic—none0/10

   New Relic's core platform is closed-source SaaS; evidence only shows an open-source CLI (newrelic-cli) and agents, not the product's source under an open license, and no evidence of the core platform's source being available.

   • [github] “Entity Search: Search for entities across all your New Relic accounts”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [community] “They don’t care that much about open source agent, but they do care about where the server is... there is still a huge underserved market fo…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb is a closed, commercial SaaS observability platform; no evidence indicates its source code is published under any open license (only OpenAPI spec download and OSS OpenTelemetry standard support are mentioned, not the product's own source).

   ⚑ flag

4. ai-native userSelf-host the core product

   weight 3 · round drawn
   New Relic—none0/10

   New Relic is documented throughout as a SaaS platform (agents forwarding data to New Relic's cloud, hosted dashboards, cloud-based NRQL/NerdGraph APIs) with no mention of a self-hostable core product. A community comment explicitly notes New Relic doesn't offer on-premise deployment ('they do care about where the server is... huge underserved market for on-premise APMs'), confirming the absence.

   • [community] “They don’t care that much about open source agent, but they do care about where the server is... there is still a huge underserved market fo…”
   • [claimed-docs] “100 GB of free data ingest/month”
   • [claimed-docs] “Sign up for free access with our perpetual free tier and 100 GB/month of data ingest included.”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb is a SaaS observability platform with a free-tier pricing page and no evidence of a self-hostable/on-prem deployment option; all evidence points to hosted cloud service usage only.

   • [claimed-docs] “Our introductory plan, free forever.”
   • [claimed-docs] “Event Volume: Up to 20M per month Metrics Data Points: Up to 100M per month”

   proof ↗⚑ flag

Otel

1. developerSend telemetry directly over OTLP with first-class OpenTelemetry support

   weight 3 · round to Honeycomb
   New Relic~partialclaimed6/10

   New Relic docs explicitly promote OpenTelemetry instrumentation for platform-agnostic observability, confirming OTLP/OTel support as a first-class ingestion path alongside native agents. However, evidence lacks technical detail on OTLP endpoint configuration, semantic-convention mapping fidelity, or independent hands-on validation of OTLP data quality. Missing for 10: detailed OTLP endpoint/setup docs, community/hands-on confirmation of OTLP ingestion working well, and specifics on semantic convention compatibility.

   • [claimed-docs] “Instrument your service with OpenTelemetry and gain platform-agnostic observability.”

   proof ↗⚑ flag

   Honeycomb✓fullclaimed9/10

   Docs explicitly confirm Honeycomb natively ingests OTLP over gRPC, HTTP/protobuf, and HTTP/JSON, with first-class OpenTelemetry instrumentation guidance for traces, logs, and metrics. Missing for 10: independent hands-on verification of OTLP ingestion beyond vendor docs.

   • [claimed-docs] “Instrument your applications with OpenTelemetry, the open-source standard for collecting telemetry, and send traces, logs, and metrics to Ho…”
   • [claimed-docs] “Honeycomb supports receiving telemetry data via OpenTelemetry’s native protocol, OTLP, over gRPC, HTTP/protobuf, and HTTP/JSON.”
   • [claimed-docs] “Honeycomb supports OpenTelemetry... send traces, logs, and metrics to Honeycomb.”
   • [claimed-docs] “If your application is already instrumented with OpenTelemetry, you can send OpenTelemetry Protocol (OTLP) data directly to Honeycomb.”
   • [claimed-docs] “This example creates a span, adds context as attributes, and closes it when the work is done”

   proof ↗⚑ flag

2. sreInstrument once with open standards and switch backends without re-instrumenting my code

   weight 2 · round to Honeycomb
   New Relic~partialclaimed4/10

   New Relic docs confirm OpenTelemetry support ('Instrument your service with OpenTelemetry and gain platform-agnostic observability'), which addresses the open-standards instrumentation ask, but the evidence pack otherwise emphasizes proprietary APM agents, NRQL, and NerdGraph as the primary ingestion/analysis path, with no detail on OTLP endpoint configuration, semantic-convention parity, or documented ease of switching backends without re-instrumenting. Missing for 10: detailed OTel onboarding/ingestion docs, evidence of vendor-neutral backend portability, and independent/hands-on confirmation that OTel-instrumented apps get full feature parity vs proprietary agents.

   • [claimed-docs] “Instrument your service with OpenTelemetry and gain platform-agnostic observability.”
   • [claimed-docs] “Monitor application performance across your stack. We’ve got agents for Go, Java, .NET, Node.js, PHP, Python, and Ruby.”
   • [claimed-docs] “Monitor application performance across your stack. We've got agents for Go, Java, .NET, Node.js, PHP, Python, and Ruby.”

   proof ↗⚑ flag

   Honeycomb✓fullclaimed8/10

   Honeycomb explicitly documents OpenTelemetry as its recommended instrumentation path, supporting native OTLP over gRPC/HTTP, meaning apps instrumented with vendor-neutral OTel SDKs can send data to Honeycomb without custom re-instrumentation, and could similarly point that same OTel pipeline at another OTLP-compatible backend. Missing for 10: explicit documentation/case study demonstrating a customer switching backends while reusing the same instrumentation, and independent confirmation of zero vendor lock-in beyond OTLP ingestion.

   • [claimed-docs] “Instrument your applications with OpenTelemetry, the open-source standard for collecting telemetry, and send traces, logs, and metrics to Ho…”
   • [claimed-docs] “Honeycomb supports receiving telemetry data via OpenTelemetry’s native protocol, OTLP, over gRPC, HTTP/protobuf, and HTTP/JSON.”
   • [claimed-docs] “Honeycomb supports OpenTelemetry... send traces, logs, and metrics to Honeycomb.”
   • [claimed-docs] “If your application is already instrumented with OpenTelemetry, you can send OpenTelemetry Protocol (OTLP) data directly to Honeycomb.”
   • [claimed-docs] “This example creates a span, adds context as attributes, and closes it when the work is done”

   proof ↗⚑ flag

1. ai-native userChoose where my data is stored (region/residency)

   weight 2 · round drawn
   New Relic—none0/10

   The evidence pack contains no mention of region selection, EU/US data center choice, or data residency controls; community comments only note latency issues from non-US access, not an actual data-locality control. Since this is a plausible axis for a SaaS observability platform, absence of evidence yields 'none'.

   • [community] “New Relic is especially slow because of the added latency of the trans-oceanic network hop [for non-US users]. The developers generally avoi…”
   • [community] “They don’t care that much about open source agent, but they do care about where the server is... there is still a huge underserved market fo…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   No evidence pack item mentions data region/residency options, EU/US storage choices, or data localization controls for Honeycomb; the pack only covers ingestion, querying, alerting, MCP, and API features.

   ⚑ flag

2. ai-native userPrevent my data from being used to train AI models

   weight 3 · round drawn
   New Relic—none0/10

   No evidence in the pack addresses data usage for AI model training or an opt-out/data-privacy control for that purpose; the AI-related docs describe New Relic AI assistant features and MCP server, not training data policies. Missing for 10: any privacy policy, data processing agreement, or opt-out mechanism specifically regarding AI training use of customer data.

   ⚑ flag

   Honeycomb—none0/10

   The evidence pack contains no mention of AI/ML training data usage policies, opt-out mechanisms, or data privacy commitments regarding AI model training for Honeycomb's product data.

   ⚑ flag

3. ai-native userControl data retention and deletion

   weight 2 · round drawn
   New Relic—none0/10

   No evidence pack items address data retention policies, deletion controls, or privacy/compliance configuration for AI-native users; the pack covers monitoring, dashboards, alerts, and MCP/AI query features but nothing on retention/deletion controls.

   ⚑ flag

   Honeycomb—none0/10

   Evidence only mentions that Honeycomb tracks data 'past your retention period' for SLOs, implying a retention policy exists, but there is no documentation showing users can configure retention periods, request data deletion, or otherwise control data lifecycle as part of an AI-native privacy posture.

   • [claimed-docs] “Honeycomb will track SLO values past your retention period, but will display these for only the Budget Burndown and Historical Compliance gr…”

   proof ↗⚑ flag

4. ai-native userOpt out of telemetry and usage tracking

   weight 2 · round drawn
   New Relic—none0/10

   No documentation addresses whether New Relic's AI features (New Relic AI assistant, MCP server) track usage/telemetry or provide an opt-out; the only related evidence (community tip on disabling APM log-forwarding via env vars) concerns application log collection, not AI-native usage tracking.

   • [community] “Setting NEW_RELIC_APPLICATION_LOGGING_ENABLED=false and NEW_RELIC_APPLICATION_LOGGING_FORWARDING_ENABLED=false disables the newly-default au…”
   • [claimed-docs] “It allows you to ask questions, troubleshoot issues, and explore telemetry data using plain language.”
   • [claimed-docs] “This update provides a standardized, scalable way to connect your favorite AI development tools directly to the rich context of the New Reli…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   No evidence pack content addresses telemetry opt-out or usage-tracking controls for AI-native users; Honeycomb's docs focus on its role as a data-collection/observability platform, not on disabling tracking of its own product usage.

   ⚑ flag

Analysis

1. developerGroup and filter by high-cardinality fields (user id, request id) without pre-aggregating or defining indexes first

   weight 2 · round to Honeycomb
   New Relic~partialcommunity3/10

   NRQL is described as a flexible ad-hoc query language for exploring telemetry data and building charts/alerts (new-relic-docs-3, new-relic-docs-21, new-relic-docs-25), which implies some ability to filter/group on arbitrary attributes, but nothing in the evidence explicitly claims support for high-cardinality fields like user id or request id, or states that no pre-aggregation/indexing is required. A community report even notes 'constant problems with metric names exploding' (new-relic-comm-11), hinting at cardinality friction in practice. missing for 10: explicit documentation of high-cardinality attribute support, confirmation that NRQL/NRDB requires no schema/index definition for arbitrary fields, and hands-on evidence of filtering by user id/request id at scale.

   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [community] “I've long preferred New Relic as well - it always seemed the most natural to use; the information that was important was easy to find... I'v…”

   proof ↗⚑ flag

   Honeycomb✓fullcommunity7/10

   Honeycomb's Query Builder supports GROUP BY, WHERE, and other clauses directly against raw event/trace data (including relational span prefixes like root./parent./child.) without requiring predefined indexes, and BubbleUp is referenced as an ad-hoc outlier/grouping tool — this is the core high-cardinality investigation workflow Honeycomb is built around, corroborated by community use for investigations. missing for 10: explicit documentation stating 'no pre-aggregation/no indexing required' or benchmarks/independent tests specifically demonstrating high-cardinality field grouping (e.g., by user_id) at scale.

   • [claimed-docs] “Query Builder allows you to construct queries against your data to produce results for investigation and further exploration.”
   • [claimed-docs] “A query in Honeycomb consists of up to six clauses”
   • [claimed-docs] “Relational field span prefixes follow your trace structure and include: root., parent., child., anyX.”
   • [claimed-docs] “A query in Honeycomb consists of up to six clauses: SELECT... WHERE... GROUP BY... ORDER BY... LIMIT... HAVING”
   • [claimed-docs] “BubbleUp ![Checkmark]”
   • [community] “I like Honeycomb a lot and we're dependent on it for parts of our orchestrator. It's great; it accelerates investigations. But even with Hon…”

   proof ↗⚑ flag

Errors

1. developerSee application errors grouped into issues with stack traces, release tracking, and regression detection

   weight 2 · round drawn
   New Relic—none0/10

   The evidence pack covers APM agent instrumentation, NRQL querying, dashboards, and deployment markers, but nowhere mentions error grouping into issues, stack trace capture, or regression detection — the core elements of this story. Missing for 10: any documentation of an 'errors inbox' or issue-grouping UI, stack trace capture, or regression detection tied to deployments/releases.

   • [claimed-docs] “Monitor application performance across your stack. We've got agents for Go, Java, .NET, Node.js, PHP, Python, and Ruby.”
   • [github] “Deployment Markers: Easily record an APM Application deployment within New Relic.”
   • [claimed-docs] “Our log management solution provides deeper visibility into application and infrastructure performance data (events, errors, traces, and mor…”

   proof ↗⚑ flag

   Honeycomb—none0/10

   Honeycomb's evidence focuses on traces/spans, OTel ingestion, ad-hoc querying, Boards, Triggers/SLOs, and BubbleUp outlier detection — none of the evidence describes grouping errors into 'issues,' capturing stack traces, release/version tracking, or automated regression detection, which are the hallmarks of dedicated error-tracking tools rather than Honeycomb's trace-analytics model.

   • [claimed-docs] “Query Builder allows you to construct queries against your data to produce results for investigation and further exploration.”
   • [claimed-docs] “This example creates a span, adds context as attributes, and closes it when the work is done”
   • [claimed-docs] “A query in Honeycomb consists of up to six clauses: SELECT... WHERE... GROUP BY... ORDER BY... LIMIT... HAVING”

   proof ↗⚑ flag

Query language

1. developerAnalyze telemetry ad hoc with a documented query language

   weight 3 · round drawn
   New Relic✓fullclaimed9/10

   NRQL is a well-documented query language for ad hoc analysis (creating charts, troubleshooting, alerts), with an API explorer for experimentation via NerdGraph, and even an AI assistant to help construct queries. Missing for 10: independent hands-on validation of NRQL usability beyond vendor docs (community comments focus on cost/UI issues rather than the query language itself).

   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”
   • [claimed-docs] “NRQL can help you: Create a new chart ... Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “Set up NRQL-based alerts (our primary and most powerful type of alert)”
   • [claimed-docs] “NerdGraph is our GraphQL-format API that lets you query New Relic data and configure some New Relic features.”
   • [claimed-docs] “An easy way to experiment with the NerdGraph API is to use the NerdGraph API explorer.”
   • [claimed-docs] “If you're struggling to get a query to work, ask our AI assistant to help you.”

   proof ↗⚑ flag

   Honeycomb✓fullcommunity9/10

   Honeycomb's Query Builder is documented with a formal structured query language (SELECT/WHERE/GROUP BY/ORDER BY/LIMIT/HAVING clauses, relational span prefixes) for ad hoc exploration of telemetry, plus Boards/BubbleUp for saved analysis and a Query Assistant for NLQ-to-query generation, with community confirmation it accelerates investigations. Missing for 10: independent hands-on benchmarking of query language expressiveness/performance at scale beyond one HN comment.

   • [claimed-docs] “Query Builder allows you to construct queries against your data to produce results for investigation and further exploration.”
   • [claimed-docs] “A query in Honeycomb consists of up to six clauses”
   • [claimed-docs] “Relational field span prefixes follow your trace structure and include: root., parent., child., anyX.”
   • [claimed-docs] “A query in Honeycomb consists of up to six clauses: SELECT... WHERE... GROUP BY... ORDER BY... LIMIT... HAVING”
   • [claimed-docs] “Relational field span prefixes follow your trace structure and include: root.- ... parent.- ... child.- ... anyX.”
   • [claimed-docs] “Query Assistant is a feature that generates Honeycomb queries based on your natural language query (NLQ) input.”
   • [claimed-docs] “we offer pre-configured Board Templates for common use cases... saving you time that would otherwise be spent building Boards manually.”
   • [community] “I like Honeycomb a lot and we're dependent on it for parts of our orchestrator. It's great; it accelerates investigations. But even with Hon…”

   proof ↗⚑ flag

Correlation

1. developerJump from a trace span to its correlated logs and metrics to debug a request end to end

   weight 2 · round to New Relic
   New Relic✓fullclaimed7/10

   New Relic's docs explicitly describe 'logs in context' where APM agents add metadata to logs enabling users to 'see logs data in various relevant places in our platform,' directly supporting jumping from traces to correlated logs, and log management ties together events, errors, and traces to reduce MTTR. Missing for 10: no independent/hands-on account specifically demonstrating the trace-span-to-log-and-metric pivot in the UI, and no detail on metrics correlation beyond logs-in-context.

   • [claimed-docs] “By default, our APM agents do two things: add metadata to your logs, which gives you logs in context (ability to see logs data in various re…”
   • [claimed-docs] “Use our APM agents to report logs. ... it lets you easily report application logs, with no additional third-party solutions required.”
   • [claimed-docs] “Our log management solution provides deeper visibility into application and infrastructure performance data (events, errors, traces, and mor…”
   • [claimed-docs] “Get visibility into hosts, containers, providers, network, and infrastructure.”

   proof ↗⚑ flag

   Honeycomb~partialcommunity5/10

   Honeycomb ingests traces, logs, and metrics together via OpenTelemetry (docs-1, docs-14, docs-15) and its Query Builder can traverse trace structure with root/parent/child relational prefixes (docs-17, docs-25), suggesting some cross-signal correlation within a single dataset. However, there is no explicit documentation of a UI action to 'jump' from a specific span directly to its correlated logs/metrics view, and community evidence (honeycomb-comm-2) indicates practitioners still manually cross-reference separate tools (Honeycomb traces, OpenSearch logs, Prometheus metrics) during incidents, suggesting the seamless single-pane correlation described in the story isn't fully realized. Missing for 10: explicit docs/screenshots of an in-trace-view link/button to jump to correlated logs and metrics, and independent hands-on confirmation that this workflow is smooth.

   • [claimed-docs] “Instrument your applications with OpenTelemetry, the open-source standard for collecting telemetry, and send traces, logs, and metrics to Ho…”
   • [claimed-docs] “Honeycomb supports OpenTelemetry... send traces, logs, and metrics to Honeycomb.”
   • [claimed-docs] “If your application is already instrumented with OpenTelemetry, you can send OpenTelemetry Protocol (OTLP) data directly to Honeycomb.”
   • [claimed-docs] “Relational field span prefixes follow your trace structure and include: root., parent., child., anyX.”
   • [claimed-docs] “Relational field span prefixes follow your trace structure and include: root.- ... parent.- ... child.- ... anyX.”
   • [community] “One of the interesting things an agent can do that no individual telemetry tool does effectively is make deductions and integrate informatio…”

   proof ↗⚑ flag

Instrumentation

1. sreInstrument hosts, containers, Kubernetes, and cloud services through vendor-maintained agents and integrations

   weight 2 · round to New Relic
   New Relic✓fullcommunity7/10

   New Relic docs confirm vendor-maintained APM agents for major languages, infrastructure monitoring for hosts/containers/providers/network, OpenTelemetry support, Kubernetes/cloud cost visibility, and Lambda serverless monitoring—covering the breadth of hosts, containers, K8s, and cloud services the story asks for. However, community reports flag real friction (fragile in-cluster collection services, a logging-agent bug that dropped logs) suggesting operational rough edges rather than pure vendor-doc perfection. Missing for 10: independent hands-on validation of Kubernetes-specific agent reliability, dedicated K8s integration docs beyond generic infra visibility, and confirmation that in-cluster collection issues have been resolved.

   • [claimed-docs] “Monitor application performance across your stack. We've got agents for Go, Java, .NET, Node.js, PHP, Python, and Ruby.”
   • [claimed-docs] “Instrument your service with OpenTelemetry and gain platform-agnostic observability.”
   • [claimed-docs] “Gain real-time visibility into multi-cloud and Kubernetes spend.”
   • [claimed-docs] “Monitor your serverless AWS Lambda functions using both CloudWatch data and code-level instrumentation.”
   • [claimed-docs] “Get visibility into hosts, containers, providers, network, and infrastructure.”
   • [community] “currently using NR, but it's slow, expensive and the in-cluster collection services are frustratingly fragile.”
   • [community] “New Relic's logging agent had a bug that lost a good portion of our logs for a few months last year. Support were extremely lethargic about …”

   proof ↗⚑ flag

   Honeycomb~partialclaimed5/10

   Honeycomb documents broad OpenTelemetry/OTLP ingestion (gRPC, HTTP/protobuf, HTTP/JSON) as its primary instrumentation path, which covers hosts, containers, and cloud services generically via the OTel ecosystem, but the evidence pack contains no vendor-maintained Honeycomb-specific agent, Kubernetes operator, or cloud-provider integration list — it relies entirely on the OTel standard rather than first-party agents for each surface. missing for 10: a Honeycomb-branded/maintained agent or K8s integration, explicit cloud-service (AWS/GCP/Azure) integrations, and independent confirmation of ease of setup across these environments.

   • [claimed-docs] “Instrument your applications with OpenTelemetry, the open-source standard for collecting telemetry, and send traces, logs, and metrics to Ho…”
   • [claimed-docs] “Honeycomb supports receiving telemetry data via OpenTelemetry’s native protocol, OTLP, over gRPC, HTTP/protobuf, and HTTP/JSON.”
   • [claimed-docs] “Honeycomb supports OpenTelemetry... send traces, logs, and metrics to Honeycomb.”
   • [claimed-docs] “If your application is already instrumented with OpenTelemetry, you can send OpenTelemetry Protocol (OTLP) data directly to Honeycomb.”

   proof ↗⚑ flag

Signals

1. sreCollect metrics, logs, and traces in one platform and pivot between them with shared context

   weight 3 · round to New Relic
   New Relic✓fullcommunity7/10

   New Relic's docs describe a single platform combining APM metrics (docs-1), OpenTelemetry traces (docs-4), infrastructure metrics (docs-31), and logs explicitly linked 'in context' with APM/trace data via automatic metadata tagging (docs-23, docs-29, docs-30), plus NRQL/dashboards to query and visualize across all these data types from 'anywhere in the platform' (docs-3, docs-7, docs-26). This directly matches the SRE story of unified metrics/logs/traces with cross-navigation. Missing for 10: independent/hands-on confirmation that pivoting between telemetry types is seamless in practice, and community feedback instead raises UI slowness/confusion (comm-13, comm-14) and a past logs-forwarding bug (comm-1) that temper full confidence without rising to a concrete contradiction of the unification claim itself.

   • [claimed-docs] “Monitor application performance across your stack. We've got agents for Go, Java, .NET, Node.js, PHP, Python, and Ruby.”
   • [claimed-docs] “Instrument your service with OpenTelemetry and gain platform-agnostic observability.”
   • [claimed-docs] “Use our APM agents to report logs. ... it lets you easily report application logs, with no additional third-party solutions required.”
   • [claimed-docs] “Our log management solution provides deeper visibility into application and infrastructure performance data (events, errors, traces, and mor…”
   • [claimed-docs] “By default, our APM agents do two things: add metadata to your logs, which gives you logs in context (ability to see logs data in various re…”
   • [claimed-docs] “Get visibility into hosts, containers, providers, network, and infrastructure.”
   • [claimed-docs] “NRQL can help you: Create a new chart, Answer a specific question for the purpose of troubleshooting or business analysis, Set up NRQL-based…”
   • [claimed-docs] “Through our UI you can build and share tailored user-friendly charts and visualizations.”
   • [claimed-docs] “Through our UI you can build and share tailored user-friendly charts and visualizations. Go to our dashboard capability to gather and chart …”
   • [community] “I used New Relic for the first time at my current job, and while the service it provides to our team is invaluable, the UI is slow and very …”
   • [community] “currently using NR, but it's slow, expensive and the in-cluster collection services are frustratingly fragile.”

   proof ↗⚑ flag

   Honeycomb~partialcommunity7/10

   Honeycomb documents ingesting traces, logs, and metrics all via OTLP into one platform, with a unified Query Builder (including relational span prefixes like root./parent./child.) and Boards to organize and pivot across analyses, satisfying much of the 'unified telemetry with shared context' story. However, community evidence shows at least one heavy Honeycomb user still relying on separate tools (OpenSearch for logs, Prometheus/VictoriaMetrics for metrics) alongside Honeycomb, suggesting real-world consolidation of all three signal types in one pane isn't always realized. Missing for 10: independent corroboration that logs/metrics pivoting works as seamlessly as traces, dedicated metrics-exploration UI documentation, and case studies of cross-signal correlation in practice.

   • [claimed-docs] “Instrument your applications with OpenTelemetry, the open-source standard for collecting telemetry, and send traces, logs, and metrics to Ho…”
   • [claimed-docs] “Honeycomb supports receiving telemetry data via OpenTelemetry’s native protocol, OTLP, over gRPC, HTTP/protobuf, and HTTP/JSON.”
   • [claimed-docs] “Query Builder allows you to construct queries against your data to produce results for investigation and further exploration.”
   • [claimed-docs] “Relational field span prefixes follow your trace structure and include: root., parent., child., anyX.”
   • [claimed-docs] “A Board is a workspace where you can save, organize, and share analysis components focused on a common objective.”
   • [community] “One of the interesting things an agent can do that no individual telemetry tool does effectively is make deductions and integrate informatio…”

   proof ↗⚑ flag