Warp vs Ghostty

Agent access

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

   weight 2 · round to Warp
   Warp✓fullprobed9/10

   Warp publishes a dedicated llms.txt (confirmed live via HTTP 200 probe at docs.warp.dev/llms.txt) describing its agentic docs, plus a whole family of agent-oriented doc pages under the _llms-txt/ path (code, warp-agent-cli, automation-platform, api-reference, knowledge-and-collaboration, factories), giving AI agents a structured, machine-readable entry point into Warp's documentation. Missing for 10: no independent/community confirmation that an external agent has actually consumed this llms.txt successfully.

   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.warp.dev/llms.txt # Warp > Documentation for Warp, the agentic development environment. Covers the…”
   • [claimed-docs] “Warp comes with a native code editor designed for quick, in-flow edits alongside your Agent conversations... with essentials like syntax hig…”
   • [claimed-docs] “Because it’s a regular command-line program, it works in any terminal emulator, including over SSH.”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”
   • [claimed-docs] “open the run transcript; inspect the prompt, plan, commands, logs, and output; and share the session link with teammates for review.”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “The Admin Panel gives team administrators centralized control over organization-wide Warp settings, including agent autonomy, privacy contro…”
   • [claimed-docs] “A software factory takes in requests (bug reports, feature specs, support escalations), and a coordinated fleet of agents works them into a …”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty is a terminal emulator, not a service with agent-facing docs; a direct probe confirms llms.txt and docs.md endpoints both return 404, meaning no agent-oriented docs are published for it to point at.

   • [probe] “PROBE llms.txt: HTTP 404 at https://ghostty.org/llms.txt”
   • [probe] “PROBE docs-md: HTTP 404 at https://ghostty.org/docs.md”

   proof ↗⚑ flag

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

   weight 2 · round to Warp
   Warp✓fullclaimed8/10

   Warp ships the Oz CLI explicitly documented for running/managing cloud agents 'from any terminal, script, or CI pipeline,' with API keys for headless/non-interactive auth 'ideal for CI pipelines, headless servers, and containers,' plus the standalone Warp Agent CLI that works over SSH/any terminal without the desktop app. Schedules and webhook/event triggers further support automation use cases outside interactive sessions. Missing for 10: independent/hands-on verification of actual CI pipeline runs and more detail on exit codes/output parsing for scripting.

   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “Because it’s a regular command-line program, it works in any terminal emulator, including over SSH.”
   • [claimed-docs] “The Warp Agent CLI lets you switch directories while in an agent session, and even run agents on remote machines without installing remote b…”
   • [claimed-docs] “Triggers: React to events from Slack, Linear, GitHub, or custom webhooks”
   • [claimed-docs] “Schedules: Run recurring tasks like dependency updates or dead code removal”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty is a GUI terminal emulator with GTK/macOS windowing dependencies (ghostty-comm-3, ghostty-comm-4), and a runtime probe found no remote-control/IPC interface, meaning scripts or agents cannot drive a running instance or invoke it headlessly for CI automation (ghostty-probe-rt-1). libghostty is for embedding terminal emulation in other apps, not for running Ghostty itself headlessly in CI.

   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [community] “Building from source needs specific deps: 'sudo apt install libgtk-4-dev libadwaita-1-dev git', otherwise you get 'unable to spawn glib-comp…”
   • [community] “It also seems to need a fairly new gtk4. At least the version in Ubuntu 22.04 is too old.”

   proof ↗⚑ flag

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

   weight 3 · round to Warp
   Warp✓fullclaimed9/10

   Warp has documented, first-party support for connecting MCP servers to extend its agents ('MCP servers extend Warp's local agents... acting as plugins for Warp'), plus explicit CLI support (Oz CLI) to 'connect MCP servers like GitHub and Linear' and agent profiles to control MCP server usage. This is corroborated across multiple doc pages describing MCP as a first-class integration point for tool use. Missing for 10: independent/hands-on community confirmation of MCP server usage specifically (community evidence covers other topics, not MCP directly).

   • [claimed-docs] “Connect Warp AI to any tool via the Model Context Protocol. Pull context from databases, APIs, docs, and internal services so the AI underst…”
   • [claimed-docs] “MCP servers extend Warp’s local agents in a modular, flexible way by exposing custom tools or data sources through a standardized interface …”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”
   • [claimed-docs] “Agent Profiles - Use agent profiles to control what the agent can access, how it behaves, and where it can act, including file access, comma…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The axis applies to terminal emulators as a kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is "none", never "na". (na/none harmonized at arena bring-up — see pipeline/scripts/terminals-na-harmonize.ts.)

   ⚑ flag

4. ai-native userUse an official CLI

   weight 2 · round to Warp
   Warp✓fullclaimed8/10

   Warp ships multiple official CLIs — the standalone Warp Agent CLI usable in any terminal/SSH (warp-docs-23, warp-docs-41, warp-docs-59, warp-docs-69) and the Oz CLI for managing cloud agents in scripts/CI (warp-docs-22, warp-docs-37, warp-docs-52) — both extensively documented as first-party, agent-native tooling with API key auth for headless use (warp-docs-63). Missing for 10: independent/hands-on community corroboration specifically validating the CLI's real-world reliability (community evidence covers the desktop app, not the CLI).

   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “The Warp Agent CLI lets you switch directories while in an agent session, and even run agents on remote machines without installing remote b…”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “The Warp Agent is now available as a standalone CLI that you can use anywhere – Ghostty, iTerm 2, VS Code, built-in Windows and Mac Terminal…”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”
   • [claimed-docs] “Because it’s a regular command-line program, it works in any terminal emulator, including over SSH.”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “It creates and maintains its own pseudoterminal (PTY), which lets you and the agent execute commands and interact with running processes thr…”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty only exposes a launch-time CLI (config flags, +list-actions/+help for bindable actions) but a runtime probe explicitly found no remote-control/IPC interface, meaning an external script or AI agent cannot send commands to or read output from a running Ghostty instance — the CLI is not built for agentic automation.

   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”

   proof ↗⚑ flag

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

   weight 3 · round to Warp
   Warp✓fullprobed8/10

   Warp explicitly advertises an 'API, CLI, SDK and MCP' platform (warp-docs-42), and a live probe confirms a real OpenAPI spec is served at docs.warp.dev/openapi.json (warp-probe-2), backed by a documented CLI reference (warp-probe-3) and API-key based headless auth for the Oz CLI (warp-docs-63) plus scriptable agent control (warp-docs-22, warp-docs-37, warp-docs-52). Missing for 10: independent/third-party corroboration of API usage and more detailed public API reference content beyond the openapi.json existence check.

   • [claimed-docs] “API, CLI, SDK and MCP built as a platform, not a vertical product or AI teammate.”
   • [probe] “PROBE openapi: HTTP 200 at https://docs.warp.dev/openapi.json — contains "openapi" key”
   • [probe] “official CLI documented at https://docs.warp.dev/reference/cli”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”

   proof ↗⚑ flag

   Ghostty~partialprobed4/10

   Ghostty documents a macOS-only AppleScript dictionary for scripting windows/tabs/terminals/layouts and input events, and a libghostty C/Zig library for embedding terminal functionality, but these are aimed at embedding or macOS-specific automation, not a general documented public API for driving a running instance. A hands-on runtime probe found no remote-control/IPC interface — only a config-actions CLI — so an external agent/script cannot send commands to or read output from a live Ghostty session on the tested platform. Missing for 10: cross-platform scripting API, evidence AppleScript automation actually works end-to-end for external control, ability to read terminal output/state programmatically, and independent corroboration beyond vendor docs.

   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [github] “libghostty is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality”
   • [github] “`libghostty-vt` is already available and usable today for Zig and C and is compatible for macOS, Linux, Windows, and WebAssembly.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [probe] “PROBE llms.txt: HTTP 404 at https://ghostty.org/llms.txt”
   • [probe] “PROBE docs-md: HTTP 404 at https://ghostty.org/docs.md”
   • [probe] “PROBE openapi: all candidate paths 404 (https://ghostty.org/openapi.json, https://ghostty.org/swagger.json, https://ghostty.org/api/openapi.…”

   proof ↗⚑ flag

6. ai-native userBuild against official SDKs

   weight 2 · round to Ghostty
   Warp~partialprobed3/10

   Warp markets itself as 'API, CLI, SDK and MCP built as a platform' (warp-docs-42) and a live OpenAPI spec is confirmed (warp-probe-2), showing some real API surface, but there is no dedicated SDK documentation, language-specific client libraries, install instructions, or code samples anywhere in the evidence. missing for 10: dedicated SDK docs/reference, language bindings (e.g. Python/JS packages), code samples, versioning/changelog for an SDK.

   • [claimed-docs] “API, CLI, SDK and MCP built as a platform, not a vertical product or AI teammate.”
   • [probe] “PROBE openapi: HTTP 200 at https://docs.warp.dev/openapi.json — contains "openapi" key”
   • [probe] “PROBE llms.txt: HTTP 200 at https://docs.warp.dev/llms.txt # Warp > Documentation for Warp, the agentic development environment. Covers the…”

   proof ↗⚑ flag

   Ghostty~partialprobed5/10

   Ghostty does publish an official SDK — `libghostty`/`libghostty-vt`, a documented cross-platform C/Zig library for building terminal emulators or embedding terminal functionality, available across macOS/Linux/Windows/WASM. However, this is a general embedding library, not an AI-native or agent-oriented SDK, and a runtime probe found no remote-control/IPC interface, meaning agents can't programmatically drive a running Ghostty instance beyond config-time CLI actions. Missing for 10: AI-specific SDK documentation/examples, agent-facing runtime control API, and independent developer corroboration of building against libghostty for agentic use cases.

   • [github] “libghostty is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality”
   • [github] “`libghostty` is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality (suc…”
   • [github] “`libghostty-vt` is already available and usable today for Zig and C and is compatible for macOS, Linux, Windows, and WebAssembly.”
   • [claimed-docs] “libghostty provides the core terminal emulation, font handling, and rendering capabilities.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”

   proof ↗⚑ flag

7. ai-native userSubscribe to events via webhooks

   weight 2 · round to Warp
   Warp~partialclaimed4/10

   Warp's automation platform docs mention 'Triggers: React to events from Slack, Linear, GitHub, or custom webhooks' and cloud agents 'triggered by... integrations like Slack and GitHub,' indicating webhook-based triggering exists, but this describes Warp consuming/reacting to inbound webhooks to kick off agents rather than a documented outbound webhook subscription mechanism for external systems to receive Warp's own events. Missing for 10: explicit documentation of an outbound webhook/event-subscription API, payload schema, or delivery guarantees, and independent confirmation of this feature working.

   • [claimed-docs] “Triggers: React to events from Slack, Linear, GitHub, or custom webhooks”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The axis applies to terminal emulators as a kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is "none", never "na". (na/none harmonized at arena bring-up — see pipeline/scripts/terminals-na-harmonize.ts.)

   ⚑ flag

Agentic features

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

   weight 2 · round to Warp
   Warp✓fullclaimed8/10

   Warp's Automation Platform explicitly supports cloud agents that run autonomously in the background, triggered by schedules, webhooks, or integrations like Slack/GitHub/Linear (warp-docs-4, warp-docs-5, warp-docs-60), with observability/audit trails (warp-docs-53, warp-docs-61) and CLI/API control via Oz CLI for headless/CI use (warp-docs-22, warp-docs-37, warp-docs-63). missing for 10: independent/hands-on verification of autonomous background runs actually completing unattended, and more detail on failure handling/reliability at scale.

   • [claimed-docs] “Triggers: React to events from Slack, Linear, GitHub, or custom webhooks”
   • [claimed-docs] “Schedules: Run recurring tasks like dependency updates or dead code removal”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”
   • [claimed-docs] “open the run transcript; inspect the prompt, plan, commands, logs, and output; and share the session link with teammates for review.”
   • [claimed-docs] “Observability: Every run is tracked, auditable, and shareable with your team”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty is a terminal emulator, not an automation/orchestration platform; evidence shows only config, theming, and AppleScript scripting for launching windows/tabs, with a runtime probe explicitly confirming no remote-control/IPC interface exists for scripts or agents to run background automations. No evidence of scheduled tasks, background job runners, or agent-triggerable automation capabilities.

   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”

   proof ↗⚑ flag

Api quality

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

   weight 2 · round to Warp
   Warp✓fullprobed8/10

   A probe directly confirms an OpenAPI JSON spec is served at docs.warp.dev/openapi.json returning a valid 'openapi' key, and docs reference an API/CLI/SDK/MCP platform with API key management, corroborating a machine-readable spec exists. Missing for 10: no independent/third-party confirmation of spec completeness or versioning, and no docs page walking through consuming the spec.

   • [probe] “PROBE openapi: HTTP 200 at https://docs.warp.dev/openapi.json — contains "openapi" key”
   • [claimed-docs] “API, CLI, SDK and MCP built as a platform, not a vertical product or AI teammate.”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The axis applies to terminal emulators as a kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is "none", never "na". (na/none harmonized at arena bring-up — see pipeline/scripts/terminals-na-harmonize.ts.)

   ⚑ flag

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

   weight 2 · round drawn
   Warp—none0/10

   Warp exposes an API, CLI, SDK, and MCP platform (warp-docs-42, warp-probe-2 openapi.json), so API versioning is a fair axis, but no evidence pack item documents API version numbers, a changelog, or any formal deprecation policy for these interfaces.

   • [claimed-docs] “API, CLI, SDK and MCP built as a platform, not a vertical product or AI teammate.”
   • [probe] “PROBE openapi: HTTP 200 at https://docs.warp.dev/openapi.json — contains "openapi" key”
   • [probe] “official CLI documented at https://docs.warp.dev/reference/cli”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty exposes libghostty as an embeddable C/Zig API surface, which makes 'versioned API with deprecation policy' a fair question, but the evidence pack shows no mention of API versioning scheme or deprecation policy, and probes confirm no openapi/llms.txt docs describing such governance.

   • [github] “libghostty is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality”
   • [github] “`libghostty-vt` is already available and usable today for Zig and C and is compatible for macOS, Linux, Windows, and WebAssembly.”
   • [probe] “PROBE openapi: all candidate paths 404 (https://ghostty.org/openapi.json, https://ghostty.org/swagger.json, https://ghostty.org/api/openapi.…”
   • [probe] “PROBE llms.txt: HTTP 404 at https://ghostty.org/llms.txt”

   proof ↗⚑ flag

Ai agent

1. ai-native userDelegate terminal tasks to a built-in AI agent that plans and executes commands with my approval

   weight 3 · round to Warp
   Warp✓fullcommunity9/10

   Warp's built-in Agent Mode is extensively documented to plan and execute multi-step terminal tasks (writing code, running commands, controlling interactive/full-screen apps, iterating) while keeping the user in the loop via conversation view, follow-ups, and approval/steering ('watch, steer, and hand off runs'). This is corroborated across many first-party docs and independent community mentions of using it as a daily driver for AI-assisted terminal work, though some hands-on complaints exist about unrelated reliability/telemetry issues rather than the core delegate-and-approve workflow itself. Missing for 10: independent hands-on verification specifically of the approval/plan-then-execute step (vs. general agent usage), and confirmation the community complaints don't touch this exact workflow.

   • [claimed-docs] “Describe what you want to build and let Warp's built‑in agent handle the rest — writing code, running commands, and iterating until the task…”
   • [claimed-docs] “Write and refactor code across your codebase * Debug issues and fix errors * Run commands and interpret results * Plan and execute mul…”
   • [claimed-docs] “the Warp Agent can now run and control long-running or full-screen terminal applications. The agent can provide input when prompted, navigat…”
   • [claimed-docs] “By default, if you ask an AI query immediately after interacting in Agent Mode, your query is sent as a **follow-up** to the current convers…”
   • [claimed-docs] “Conversations are tied to sessions and you can run multiple Agent Mode conversations simultaneously in different windows, tabs, or panes.”
   • [claimed-docs] “any surface — watch, steer, and hand off runs from web, mobile, terminal, or IDE.”
   • [community] “Warp is great - I use it as my daily terminal. The best features are being able to edit commands, chunking the output into blocks and AI gen…”
   • [community] “My daily driver for a few months now, even after trying Rio/Ghostty/Wezterm... Favorite feature is the notification when a long running comm…”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty is a terminal emulator with no built-in AI agent, planning, or command-approval workflow anywhere in its documented feature set; evidence covers configuration, themes, rendering, and scripting via AppleScript/CLI actions only. A runtime probe explicitly found no remote-control/IPC interface that would let an external agent send or approve commands, confirming absence of this capability.

   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”

   proof ↗⚑ flag

Ai privacy

1. ai-native userUse or fully disable AI features with clear control over what session data leaves my machine, without a forced login

   weight 2 · round to Warp
   Warp!disputedcontradicted4/10

   Warp docs promise granular control (secret redaction, no training use, telemetry opt-out on paid plans, ZDR for Business/Enterprise) and note telemetry can be monitored/disabled, but this is undercut by concrete community reports: one HN thread describes Warp silently sending terminal session data to LLMs via 'Prompt Suggestions'/'Next Command' without explicit consent (warp-comm-1), and multiple independent users report being confronted with a mandatory login prompt and uninstalling immediately (warp-comm-5, warp-comm-6, warp-comm-16), even though a Warp co-founder claims the login requirement was removed (warp-comm-7). Docs also admit telemetry must stay on to use AI on the Free plan (warp-docs-33), meaning full AI use without any data leaving isn't unconditional. missing for 10: a documented explicit toggle to fully disable AI without any account/login step on the free tier, and resolution of the community claim of undisclosed data transmission.

   • [claimed-docs] “Warp does not allow OpenAI or Anthropic to use your data for model training. Our "hands off" approach also means your terminal input and out…”
   • [claimed-docs] “Monitor telemetry in real-time with Warp’s native Network Log. Opt out of telemetry at any time”
   • [claimed-docs] “Telemetry must be enabled to use AI features on the Free plan, while paid plans can opt out at any time and continue using Warp, including A…”
   • [claimed-docs] “Business and Enterprise plans are covered by our Zero Data Retention (ZDR) agreement. No AI interaction or console data is ever collected.”
   • [claimed-docs] “Read a complete list of all the telemetry events that get sent for app analytics ... Opt out of telemetry at any time”
   • [claimed-docs] “Warp unconditionally applies Secret Redaction in all AI interactions to ensure that any sensitive data is _never_ collected or sent to third…”
   • [community] “Warp silently sends terminal session data to LLMs via 'Prompt Suggestions' and 'Next Command' features without explicit consent, discovered …”
   • [community] “I tried Warp once... it showed me a login prompt. Perhaps I mistakenly downloaded a login prompt app instead of a terminal app? I promptly u…”
   • [community] “I remember seeing this terminal emulator, deciding to give it a shot, and being asked to sign in. I have never uninstalled a program faster …”
   • [community] “Warp co-founder: fair criticism, and this is the reason we removed the login requirement... the point of the login is that we have features …”
   • [community] “They have also innovated by requiring login (into a console!!)... Today they have heaps of telemetry, so a privacy nightmare. This is a term…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The evidence pack contains no mention of any AI features, AI settings, session-data controls, or login requirements for Ghostty — it is documented purely as a native terminal emulator with configuration, theming, and libghostty embedding capabilities. Since terminal emulators (e.g., AI-augmented ones) can plausibly ship such features, the axis is applicable, but no evidence shows Ghostty offers or disables AI functionality or addresses data-egress/login controls.

   ⚑ flag

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

   weight 3 · round to Warp
   Warp✓fullclaimed8/10

   Warp's Automation Platform explicitly supports Triggers reacting to events from Slack, Linear, GitHub, or custom webhooks, plus Schedules for recurring tasks, and cloud agents described as 'triggered by system events, schedules, or integrations like Slack and GitHub' with full observability/audit trails. This directly matches the story of defining rules that trigger automatic actions on events. Missing for 10: independent/hands-on verification of the trigger system in practice (community evidence focuses on core terminal/agent issues, not the automation/trigger feature specifically).

   • [claimed-docs] “Triggers: React to events from Slack, Linear, GitHub, or custom webhooks”
   • [claimed-docs] “Schedules: Run recurring tasks like dependency updates or dead code removal”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”
   • [claimed-docs] “open the run transcript; inspect the prompt, plan, commands, logs, and output; and share the session link with teammates for review.”
   • [claimed-docs] “Observability: Every run is tracked, auditable, and shareable with your team”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty offers fixed automatic behaviors (auto-detect password prompts, dark/light theme switching) and on-demand AppleScript scripting, but there is no evidence of a general user-definable rule/trigger system that fires actions on arbitrary terminal events, and a runtime probe explicitly found no remote-control/IPC surface for such automation.

   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Secure Keyboard Entry: Automatically detect password prompts or manually enable secure keyboard entry to protect passwords from other proces…”
   • [claimed-docs] “Themes can be switched automatically based on system dark/light mode.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”

   proof ↗⚑ flag

2. ai-native userSchedule recurring jobs or workflows

   weight 2 · round to Warp
   Warp✓fullclaimed8/10

   Warp documents explicit 'Schedules' functionality for cloud agents ('Run recurring tasks like dependency updates or dead code removal') alongside 'Triggers' for event-based automation, plus cloud agents described as 'triggered by system events, schedules, or integrations' and managed via the Oz CLI for scripting/CI. This directly matches the story of scheduling recurring jobs/workflows. Missing for 10: independent/hands-on confirmation of the scheduling feature working reliably, and more detail on scheduling configuration (cron-like syntax, frequency options).

   • [claimed-docs] “Triggers: React to events from Slack, Linear, GitHub, or custom webhooks”
   • [claimed-docs] “Schedules: Run recurring tasks like dependency updates or dead code removal”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The axis applies to terminal emulators as a kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is "none", never "na". (na/none harmonized at arena bring-up — see pipeline/scripts/terminals-na-harmonize.ts.)

   ⚑ flag

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

   weight 1 · round to Warp
   Warp~partialclaimed4/10

   Warp provides strong review/observability for automations (run transcripts showing prompt, plan, commands, logs, shareable for team review) and lets you save/share workflows via Warp Drive, but there is no documented versioning or rollback mechanism for automations (e.g., no version history, diffing, or revert capability for schedules/triggers/workflows). missing for 10: explicit version history for saved automations, ability to roll back to a prior automation version, changelog/diff view for automation edits.

   • [claimed-docs] “Warp lets you save your most important workflows in Warp Drive so you can run them on‑demand or share with your team.”
   • [claimed-docs] “Observability: Every run is tracked, auditable, and shareable with your team”
   • [claimed-docs] “open the run transcript; inspect the prompt, plan, commands, logs, and output; and share the session link with teammates for review.”
   • [claimed-docs] “Warp lets you save your most important workflows in Warp Drive so you can run them on‑demand or share with your team. Let Warp AI handle the…”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The axis applies to terminal emulators as a kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is "none", never "na". (na/none harmonized at arena bring-up — see pipeline/scripts/terminals-na-harmonize.ts.)

   ⚑ flag

Dotfiles

1. power-userKeep the terminal's full configuration as plain-text files in my dotfiles and sync it across machines

   weight 3 · round to Ghostty
   Warp~partialclaimed4/10

   Warp's open-source announcement mentions shipping a new settings file explicitly for 'programmatic control over settings and easy portability between devices' (warp-docs-40), which speaks directly to the dotfiles/sync use case, but this is a single passing mention with no documentation of its format, scope, or actual dotfiles-style workflow. Other config-adjacent features (theme OS-sync, session restoration) are separate app-state features, not plain-text portable config. Missing for 10: dedicated docs on the settings file format/location, confirmation it covers full app config (keybindings, workflows, themes), and independent/community corroboration of dotfiles-style syncing across machines.

   • [claimed-docs] “we are shipping a (long-overdue) settings file so that users and agents get programmatic control over settings and easy portability between …”
   • [claimed-docs] “Warp supports synchronizing your theme with the OS’s light and dark themes.”
   • [claimed-docs] “Restore your windows, tabs, panes, and recent Blocks automatically when you relaunch Warp.”

   proof ↗⚑ flag

   Ghostty✓fullclaimed8/10

   Ghostty's config is a plain-text file (config-file key), supports splitting into multiple files for modular dotfiles, every key works as a CLI flag, and it can be reloaded live without restart, all of which make it straightforward to keep in a dotfiles repo and sync across machines. missing for 10: no explicit documentation or community confirmation of cross-platform (macOS/Linux/Windows) path consistency for the config file, and no first-party dotfiles-sync tooling or guide is mentioned.

   • [claimed-docs] “You can split your configuration into multiple files by using the `config-file` key in your configuration file.”
   • [claimed-docs] “The configuration can be reloaded at runtime by pressing `ctrl+shift+,` (Linux) or `cmd+shift+,` (macOS).”
   • [claimed-docs] “The configuration can be reloaded at runtime by pressing ctrl+shift+, (Linux) or cmd+shift+, (macOS).”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”
   • [claimed-docs] “Ghostty supports hundreds of configuration options to make it look and behave exactly how you want.”

   proof ↗⚑ flag

Fonts

1. developerUse programming fonts with ligatures, fallback fonts, and fine-grained font tuning

   weight 2 · round to Ghostty
   Warp—none0/10

   No evidence in the pack mentions font configuration, ligatures, fallback fonts, or font tuning of any kind; documentation focuses on themes, appearance sync, and agent features, not typography settings.

   ⚑ flag

   Ghostty✓fullcommunity9/10

   Ghostty's docs explicitly cover ligature rendering (ghostty-docs-22/33), toggling programming ligatures via -calt (ghostty-docs-7), specifying font-features applied across all fonts (ghostty-docs-26), and configurable fallback fonts for missing codepoints (ghostty-docs-38), plus built-in nerd fonts and a default JetBrains Mono font (ghostty-docs-28). Community feedback corroborates deep, granular config tuning (minimum-contrast, selection colors) as a strength (ghostty-comm-5, ghostty-comm-6). Missing for 10: independent hands-on verification specifically of ligature rendering quality/fallback behavior beyond docs claims.

   • [claimed-docs] “You can use fonts that have ligatures and Ghostty will render them correctly.”
   • [claimed-docs] “You can use fonts that have ligatures and Ghostty will render them correctly. You can also specify specific font features to enable or disab…”
   • [claimed-docs] “To disable programming ligatures, use `-calt` since this is the typical feature name for programming ligatures.”
   • [claimed-docs] “The font feature will apply to all fonts rendered by Ghostty.”
   • [claimed-docs] “This configuration can be repeated multiple times to specify preferred fallback fonts when the requested codepoint is not available in the p…”
   • [claimed-docs] “Ghostty is designed to work out of the box with no configuration for most users. Ghostty has sensible defaults, embeds a default font (JetBr…”
   • [community] “Thanks for the minimum-contrast option (ghostty.org/docs/config/reference#minimum-contrast) - sick of tools with unreadable dark-on-black te…”
   • [community] “selection-foreground and selection-background config to set highlight colors makes a massive difference, versus the translucent selection on…”

   proof ↗⚑ flag

Theming

1. power-userApply color schemes and themes, including automatic light/dark mode switching

   weight 2 · round to Ghostty
   Warp✓fullclaimed8/10

   Warp's docs explicitly support custom themes, auto-generating themes from a background image, and syncing theme with OS light/dark mode (warp-docs-19, warp-docs-20), directly matching the story. Missing for 10: independent/community corroboration of theming reliability and more detail on breadth of built-in theme palette.

   • [claimed-docs] “Automatically create new themes based on a background image.”
   • [claimed-docs] “Warp supports synchronizing your theme with the OS’s light and dark themes.”

   proof ↗⚑ flag

   Ghostty✓fullcommunity9/10

   Docs explicitly confirm hundreds of built-in themes selectable via config, custom theme authoring, and automatic switching based on system light/dark mode, plus related fine-grained color config (minimum-contrast, selection colors) corroborated by community users praising these theming features. missing for 10: no independent hands-on demo/screenshot verifying auto light/dark switching in practice beyond docs claims.

   • [claimed-docs] “Themes: Ghostty ships with hundreds of themes that can be selected with a single line of configuration. Themes can be switched automatically…”
   • [claimed-docs] “Themes can be switched automatically based on system dark/light mode.”
   • [claimed-docs] “Ghostty ships with hundreds of built-in themes, supports different themes for light and dark mode, and more.”
   • [claimed-docs] “Users can author their own themes.”
   • [community] “Thanks for the minimum-contrast option (ghostty.org/docs/config/reference#minimum-contrast) - sick of tools with unreadable dark-on-black te…”
   • [community] “selection-foreground and selection-background config to set highlight colors makes a massive difference, versus the translucent selection on…”

   proof ↗⚑ flag

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

   weight 2 · round to Warp
   Warp~partialprobed6/10

   Warp explicitly positions itself as 'API, CLI, SDK and MCP built as a platform,' with the Oz CLI enabling agent creation, monitoring, MCP connections, and authentication without the desktop app, and a live OpenAPI spec — showing substantial API/CLI parity for agent workflows. But the story demands full UI-equivalent capability, and evidence doesn't confirm parity for many UI-only features (theming, session restoration, conversation UI, code editor) nor is there independent verification that the API surface is fully complete. missing for 10: independent confirmation of full feature parity, evidence that UI-specific features (themes, editor, session restore) are also API-accessible

   • [claimed-docs] “API, CLI, SDK and MCP built as a platform, not a vertical product or AI teammate.”
   • [claimed-docs] “any surface — watch, steer, and hand off runs from web, mobile, terminal, or IDE.”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “Agent Profiles - Use agent profiles to control what the agent can access, how it behaves, and where it can act, including file access, comma…”
   • [probe] “PROBE openapi: HTTP 200 at https://docs.warp.dev/openapi.json — contains "openapi" key”
   • [probe] “official CLI documented at https://docs.warp.dev/reference/cli”

   proof ↗⚑ flag

   Ghostty!disputedcontradicted3/10

   Ghostty documents AppleScript automation for scripting windows/tabs/terminals (macOS only) and exposes config keys as CLI flags, but a hands-on runtime probe found no remote-control/IPC interface: the CLI only supports config/action listing, and there is no way for an external script or agent to send commands to or read output from a running Ghostty instance. This directly contradicts the idea of full UI/API parity for an AI-native user. missing for 10: a documented remote-control/scripting API beyond macOS AppleScript, ability to read terminal state/output programmatically, cross-platform (Linux/Windows) automation parity.

   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”

   proof ↗⚑ flag

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

   weight 3 · round to Ghostty
   Warp~partialclaimed3/10

   Warp documents a settings file for 'programmatic control over settings and easy portability between devices' and open-sources its client code, but there's no documented feature for exporting conversation history, Warp Drive workflows, or agent run data in open formats for full user data portability. missing for 10: explicit full data export tooling, open-format export of conversation/session history, Warp Drive workflow export, independent confirmation users can fully migrate away with their data.

   • [claimed-docs] “we are shipping a (long-overdue) settings file so that users and agents get programmatic control over settings and easy portability between …”
   • [claimed-docs] “Read and audit Warp’s client source code at warpdotdev/warp, open source under AGPL v3”
   • [claimed-docs] “Warp’s source code is now available at github.com/warpdotdev/warp with an AGPL license.”

   proof ↗⚑ flag

   Ghostty~partialprobed4/10

   Ghostty's configuration lives in plain, human-readable text files that can be split, version-controlled, and passed as CLI flags, which gives a basic form of open, portable settings data — but there is no explicit 'export my data' feature, session/history export, or documented data-portability guarantee (and probes for llms.txt/docs export endpoints return 404). missing for 10: an explicit export/backup mechanism for data beyond config (e.g., scrollback, session state), documentation framing this as a data-ownership/exit story, and independent confirmation that users rely on this for full data portability.

   • [claimed-docs] “Ghostty supports hundreds of configuration options to make it look and behave exactly how you want.”
   • [claimed-docs] “You can split your configuration into multiple files by using the `config-file` key in your configuration file.”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”
   • [probe] “PROBE llms.txt: HTTP 404 at https://ghostty.org/llms.txt”
   • [probe] “PROBE docs-md: HTTP 404 at https://ghostty.org/docs.md”

   proof ↗⚑ flag

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

   weight 2 · round to Warp
   Warp✓fullclaimed8/10

   Warp explicitly published its client source code on GitHub under the AGPL v3 open-source license, documented in both its privacy docs and a dedicated blog post announcing the move to open source. missing for 10: independent/third-party confirmation of the full scope of open-sourced code (e.g., whether cloud/server-side agent components are also included) and any hands-on community audit of the repository beyond the vendor's own announcement.

   • [claimed-docs] “Read and audit Warp’s client source code at warpdotdev/warp, open source under AGPL v3”
   • [claimed-docs] “Warp’s source code is now available at github.com/warpdotdev/warp with an AGPL license.”
   • [claimed-docs] “Read and audit Warp’s client source code at `warpdotdev/warp`, open source under AGPL v3”

   proof ↗⚑ flag

   Ghostty~partialclaimed6/10

   Ghostty's source is hosted publicly on GitHub and the evidence confirms a public repo plus the libghostty/libghostty-vt libraries usable by developers, implying an open-source, readable codebase. However, no evidence item explicitly names or links the license text (e.g., MIT/Apache) confirming the terms under which the source can be read/used. Missing for 10: explicit license identification/citation, confirmation of license permissiveness, and any independent commentary on licensing terms.

   • [github] “Anyone can use `libghostty` to build a terminal emulator or embed a terminal into their own applications.”
   • [github] “libghostty is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality”
   • [github] “`libghostty` is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality (suc…”
   • [github] “`libghostty-vt` is already available and usable today for Zig and C and is compatible for macOS, Linux, Windows, and WebAssembly.”

   proof ↗⚑ flag

4. ai-native userSelf-host the core product

   weight 3 · round to Warp
   Warp~partialcommunity4/10

   Warp's client is open-sourced under AGPL v3 (warpdotdev/warp) so the terminal client itself can be inspected and self-hosted/built from source, and cloud agents are described as able to run 'on Warp's cloud infrastructure or your own,' suggesting some self-hosting of agent execution. However, there's no documented path to self-host the full core product—AI model routing, backend services, and account/login (required per warp-comm-6/7) remain tied to Warp's proprietary cloud, and no deployment/on-prem guide is evidenced. Missing for 10: a documented self-hosted backend/server deployment, evidence that AI/model-routing works without Warp's cloud, and independent confirmation of a working self-hosted setup.

   • [claimed-docs] “Read and audit Warp’s client source code at warpdotdev/warp, open source under AGPL v3”
   • [claimed-docs] “Warp’s source code is now available at github.com/warpdotdev/warp with an AGPL license.”
   • [claimed-docs] “Cloud agents are autonomous, background agents that run on Warp’s cloud infrastructure or your own, triggered by system events, schedules, o…”
   • [community] “Warp co-founder: fair criticism, and this is the reason we removed the login requirement... the point of the login is that we have features …”

   proof ↗⚑ flag

   Ghostty—none0/10

   The axis applies to terminal emulators as a kind (peer products hold positive or none verdicts on this story), so lack of evidence for an applicable capability is "none", never "na". (na/none harmonized at arena bring-up — see pipeline/scripts/terminals-na-harmonize.ts.)

   ⚑ flag

Multiplexing

1. developerSplit windows into panes and organize work across tabs without an external multiplexer

   weight 3 · round to Ghostty
   Warp~partialclaimed6/10

   Docs confirm panes exist alongside tabs and windows as first-class session units (agent conversations run in "windows, tabs, or panes" and session restoration recreates "windows, tabs, and panes"), showing built-in multiplexing without needing tmux. However the pack lacks any detail on pane-splitting mechanics, keyboard shortcuts, or layout management, and no independent/hands-on account discusses pane usage specifically. Missing for 10: dedicated documentation on split/pane creation and navigation, independent user confirmation of pane workflow.

   • [claimed-docs] “Conversations are tied to sessions and you can run multiple Agent Mode conversations simultaneously in different windows, tabs, or panes.”
   • [claimed-docs] “Restore your windows, tabs, panes, and recent Blocks automatically when you relaunch Warp.”
   • [claimed-docs] “Continue previous threads with follow-ups, manage conversation history, attach context from blocks and files, and run multiple conversations…”

   proof ↗⚑ flag

   Ghostty✓fullclaimed8/10

   Ghostty's official docs explicitly describe native support for multiple windows, each with its own tabs and splits, directly addressing the story of organizing work without an external multiplexer (ghostty-docs-32, ghostty-docs-40). Community discussion corroborates general daily-use satisfaction but does not specifically validate the pane/tab experience in depth. Missing for 10: independent hands-on validation specifically of the splits/tabs workflow (community comments focus on other features like search or scrollback), and no mention of pane resizing/navigation keybindings specifics.

   • [claimed-docs] “Ghostty supports multiple windows, each with its own tabs and splits. These are all rendered using native UI components.”
   • [claimed-docs] “Windows, tabs, and splits: Ghostty supports multiple windows, each with its own tabs and splits.”
   • [claimed-docs] “Ghostty supports flexible, custom keybindings through the keybind configuration option.”

   proof ↗⚑ flag

Quick access

1. power-userSummon the terminal instantly with a global hotkey or quake-style dropdown window

   weight 1 · round to Ghostty
   Warp—none0/10

   No evidence in the pack mentions a global hotkey or quake-style dropdown/instant-summon window feature; Warp's docs focus on session restoration, panes/tabs, and AI agent features, not a system-wide invocation shortcut.

   ⚑ flag

   Ghostty✓fullcommunity8/10

   Ghostty documents a dedicated Quick Terminal feature that animates down for instant access without interrupting work, matching the quake-style dropdown request, and this is independently corroborated by a community member calling it 'ghostty's killer feature.' Custom global keybindings are also configurable via the keybind option. Missing for 10: explicit documentation confirming the Quick Terminal is bound to a global hotkey configurable across all platforms (evidence emphasizes macOS) and no independent hands-on walkthrough of summoning it via a custom global shortcut.

   • [claimed-docs] “Quick Terminal: Lightweight terminal that animates down below the menu bar for instant access without interrupting work.”
   • [claimed-docs] “Lightweight terminal that animates down below the menu bar for instant access without interrupting work.”
   • [community] “The quick terminal feature is ghostty's killer feature for me.”
   • [claimed-docs] “Ghostty supports flexible, custom keybindings through the keybind configuration option.”

   proof ↗⚑ flag

Sessions

1. developerRestore my windows, tabs, and working directories after a restart or crash

   weight 2 · round to Warp
   Warp✓fullclaimed8/10

   Warp's docs explicitly describe session restoration that restores windows, tabs, panes, and recent Blocks automatically on relaunch, directly matching the story. Missing for 10: independent/hands-on corroboration of restore reliability after crashes (only vendor docs cited), and no detail on working-directory restoration specifics.

   • [claimed-docs] “Restore your windows, tabs, panes, and recent Blocks automatically when you relaunch Warp.”
   • [claimed-docs] “Session restoration allows you to quickly pick up where you left off in your previous terminal session.”
   • [claimed-docs] “Sometimes, you may want to prevent a sensitive Block from being saved on your computer, or you may want to clear blocks from a machine entir…”

   proof ↗⚑ flag

   Ghostty—none0/10

   No evidence in the pack of any session/window/tab restoration feature after restart or crash; docs only describe windows/tabs/splits creation and configuration, not persistence of state. This is a plausible axis for a terminal emulator (comparable tools offer session restore), but nothing here confirms Ghostty supports it.

   ⚑ flag

Benchmarks

1. power-userSee published benchmarks or measured latency/throughput numbers backing the terminal's performance claims

   weight 2 · round to Ghostty
   Warp—none0/10

   No published benchmarks, latency, or throughput figures appear anywhere in the evidence pack; documentation focuses on features and privacy, while community posts mention performance complaints but no measured numbers. missing for 10: published benchmark data, latency/throughput measurements, any comparative performance testing.

   ⚑ flag

   Ghostty~partialcommunity4/10

   No first-party published benchmarks exist in Ghostty's docs, but independent community benchmarks provide measured latency and throughput numbers (input latency improved to ~13ms in a newer test, throughput comparable to Alacritty/Ptyxis), while an older benchmark and a hands-on user reported markedly worse input latency and memory footprint than competitors. missing for 10: official vendor-published benchmark suite, consistent/reproducible throughput-latency methodology, and resolution of the conflicting community measurements.

   • [community] “According to an 11-month-old benchmark, Ghostty had the worst input latency across all contenders; a newer benchmark from ~4 months ago show…”
   • [community] “On throughput (catting a large file), Ghostty performs very well, among the same league as Alacritty and Ptyxis.”
   • [community] “After testing Ghostty for a while: input lag is higher than xfce4-terminal, font rendering is blurrier, UI is less consistent with my deskto…”

   proof ↗⚑ flag

Rendering

1. developerRely on GPU-accelerated rendering that stays fast and responsive under heavy output

   weight 3 · round to Ghostty
   Warp—none0/10

   The evidence pack contains no first-party documentation claiming GPU-accelerated rendering or describing rendering performance under heavy output; instead, community reports describe UI glitches, broken behavior, and instability during heavy use (e.g., 'horribly broken... UI glitches', 'random failures, stopping of agents'), and one commenter notes the terminal is 'just Alacritty' with no Warp-specific performance claims. Since no vendor documentation asserts this capability, there's nothing concrete to substantiate 'full' or 'partial', and the negative community signals reinforce a 'none' verdict rather than a dispute of an explicit claim.

   • [community] “Warp is so horribly broken right now and has been for weeks. Multiple github issues on consistent issue writing file. On top of that UI glit…”
   • [community] “I've been a Pro user for several months (use < 1000 credits a month) - but I've noticed a real reduction in quality over the past month or s…”
   • [community] “Warp is a terrible product, and I have nothing nice to say about them. But look at modern terminals like Kitty or Ghostty... mouse support t…”
   • [community] “Their terminal is just Alacritty... The terminal emulator was never their selling point, the AI wrapper was.”

   proof ↗⚑ flag

   Ghostty!disputedcontradicted5/10

   Ghostty's docs/github tier claim libghostty provides core rendering performance (rendering capabilities, cross-platform GPU-based terminal library), but hands-on community reports concretely contradict smooth, fast performance under load: one user found higher input lag, blurrier fonts and 3-4x memory footprint versus a lightweight terminal, and an older independent latency benchmark showed Ghostty had the worst input latency among terminals tested — though a newer benchmark reported improved ~13ms latency and strong throughput comparable to Alacritty/Ptyxis. missing for 10: first-party performance benchmarks or documentation of GPU-rendering architecture, resolution of the reported high-memory/input-lag complaints, and consistent independent benchmark corroboration across time.

   • [claimed-docs] “libghostty provides the core terminal emulation, font handling, and rendering capabilities.”
   • [github] “libghostty is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality”
   • [github] “`libghostty` is a cross-platform, zero-dependency C and Zig library for building terminal emulators or utilizing terminal functionality (suc…”
   • [community] “After testing Ghostty for a while: input lag is higher than xfce4-terminal, font rendering is blurrier, UI is less consistent with my deskto…”
   • [community] “According to an 11-month-old benchmark, Ghostty had the worst input latency across all contenders; a newer benchmark from ~4 months ago show…”
   • [community] “On throughput (catting a large file), Ghostty performs very well, among the same league as Alacritty and Ptyxis.”
   • [community] “It feels amazing on linux. I find it very noticably better than konsole. Foot/Alacritty also feel amazing but they don't have some features …”

   proof ↗⚑ flag

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

   weight 2 · round to Warp
   Warp~partialcommunity7/10

   Warp's docs explicitly describe a telemetry opt-out and a Network Log to monitor telemetry in real time, and state paid plans can opt out at any time while still using AI. However, the Free plan requires telemetry to be enabled to use AI features, meaning opt-out isn't universal, and community commentary (warp-comm-1, warp-comm-16) raises concerns about undisclosed data sharing and 'heaps of telemetry,' which tempers full confidence though it doesn't concretely disprove the opt-out mechanism itself. missing for 10: independent verification that opt-out fully works with no residual telemetry, and free-tier parity for full opt-out.

   • [claimed-docs] “Monitor telemetry in real-time with Warp’s native Network Log. Opt out of telemetry at any time”
   • [claimed-docs] “Telemetry must be enabled to use AI features on the Free plan, while paid plans can opt out at any time and continue using Warp, including A…”
   • [claimed-docs] “Read a complete list of all the telemetry events that get sent for app analytics ... Opt out of telemetry at any time”
   • [claimed-docs] “Opt out of telemetry at any time”
   • [claimed-docs] “Monitor telemetry in real-time with Warp’s native Network Log * Opt out of telemetry at any time”
   • [community] “They have also innovated by requiring login (into a console!!)... Today they have heaps of telemetry, so a privacy nightmare. This is a term…”

   proof ↗⚑ flag

   Ghostty—none0/10

   No evidence in the pack mentions telemetry, usage tracking, or an opt-out setting for Ghostty; the docs cover configuration, features, and performance but say nothing about data collection practices. Since this is a locally-installed terminal emulator, the axis is plausible (any app could collect usage data) but unaddressed, so it defaults to none rather than na.

   ⚑ flag

Graphics

1. developerDisplay inline images and rich graphics via a documented terminal graphics protocol

   weight 2 · round to Ghostty
   Warp—none0/10

   No evidence in the pack mentions a terminal graphics protocol (e.g., Sixel, Kitty graphics protocol, iTerm2 inline images) or inline image/rich graphics rendering in Warp's terminal. Evidence only covers AI agents, MCP, telemetry, themes, and session restoration.

   ⚑ flag

   Ghostty✓fullclaimed8/10

   Ghostty's official docs explicitly document support for the Kitty graphics protocol, enabling inline image rendering directly in the terminal, and this is listed as a core feature alongside a broader reference on supported control sequences. Missing for 10: independent hands-on confirmation/testing of the graphics protocol specifically (community evidence covers other features but not image rendering) and detail on protocol coverage limits.

   • [claimed-docs] “Kitty graphics protocol: Ghostty supports the Kitty graphics protocol, which allows terminal applications to render images directly in the t…”
   • [claimed-docs] “Ghostty supports the Kitty graphics protocol, which allows terminal applications to render images directly in the terminal.”
   • [claimed-docs] “For terminal application developers, a reference on terminal concepts and supported control sequences.”

   proof ↗⚑ flag

Platforms

1. developerRun the same terminal with the same config on macOS, Linux, and Windows

   weight 2 · round to Warp
   Warp~partialcommunity5/10

   Evidence shows Warp exposes platform-specific keybindings for macOS, Windows, and Linux (docs-49), a settings file enabling config portability across devices (docs-40), and CLI/agent components that work over SSH in any terminal (docs-59), implying cross-platform availability. However, there is no explicit doc confirming identical native app experience on all three OSes, and community reports describe Warp being broken specifically under WSL/Windows-via-SSH scenarios (warp-comm-10), undercutting a seamless 'same terminal, same config' experience. missing for 10: explicit first-party confirmation of feature parity across macOS/Linux/Windows, documented config-sync mechanism beyond a settings file, and independent corroboration that the same setup truly works identically on all three platforms.

   • [claimed-docs] “Open the **Conversations menu** (`⌘+Y` on macOS, `Ctrl+Shift+Y` on Windows/Linux), select a conversation, and then enter your query.”
   • [claimed-docs] “we are shipping a (long-overdue) settings file so that users and agents get programmatic control over settings and easy portability between …”
   • [claimed-docs] “Because it’s a regular command-line program, it works in any terminal emulator, including over SSH.”
   • [community] “Warp is so horribly broken right now and has been for weeks. Multiple github issues on consistent issue writing file. On top of that UI glit…”

   proof ↗⚑ flag

   Ghostty~partialclaimed4/10

   Ghostty documents strong config portability (single config file, config-file includes, hundreds of options, CLI-flag equivalence) and confirms macOS and Linux binaries, but the evidence pack never shows an actual Ghostty terminal app for Windows — only the underlying libghostty-vt library is said to target Windows/WebAssembly, and the docs explicitly note different default keybindings on macOS vs Linux, undercutting an identical experience even across the two platforms that are supported. missing for 10: confirmed Windows terminal application/binary release, identical default keybindings across platforms, and independent testimony of using the same config file on all three OSes.

   • [claimed-docs] “Ghostty supports hundreds of configuration options to make it look and behave exactly how you want.”
   • [claimed-docs] “You can split your configuration into multiple files by using the `config-file` key in your configuration file.”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”
   • [claimed-docs] “Ghostty uses different default bindings on macOS and Linux to match the conventions of each platform.”
   • [claimed-docs] “Another part is using standard keyboard and mouse shortcuts that you're already familiar with. Ghostty uses different default bindings on ma…”
   • [github] “`libghostty-vt` is already available and usable today for Zig and C and is compatible for macOS, Linux, Windows, and WebAssembly.”
   • [claimed-docs] “A [Homebrew cask](https://formulae.brew.sh/cask/ghostty) is available and maintained by the Ghostty community.”

   proof ↗⚑ flag

Ssh

1. developerSSH to remote hosts with the terminal's own integration carrying my terminfo, shell integration, and config along

   weight 2 · round drawn
   Warp—none0/10

   Warp's docs mention running the `warp` agent CLI over SSH (warp-docs-3) but that's about launching an agent conversation, not the terminal itself carrying terminfo, shell integration, and config when SSHing into remote hosts. No evidence describes Warp's terminal propagating its shell integration/terminfo to remote hosts, and community reports (warp-comm-10, warp-comm-14) note broken functionality (file/diff explorer, code indexing) specifically over SSH connections, further indicating the core integration doesn't extend cleanly to remote sessions.

   • [claimed-docs] “Run the `warp` command to start a conversation in whichever terminal you already use, over SSH, or on a machine where Warp isn’t installed.”
   • [community] “Warp is so horribly broken right now and has been for weeks. Multiple github issues on consistent issue writing file. On top of that UI glit…”
   • [community] “Nearly everything about the experience of a terminal is better than iTerm though? Very nice having the terminal treated like a text file. It…”

   proof ↗⚑ flag

   Ghostty—none0/10

   There is no documentation in the evidence pack describing an SSH integration feature that carries terminfo, shell integration, and config to remote hosts; instead, community reports describe Ghostty's custom TERM value causing 'missing or unsuitable terminal' errors on SSH and users citing SSH breakage as a recurring blocker.

   • [community] “Using its own TERM is a deliberate design decision that can cause SSH 'missing or unsuitable terminal' errors on servers that hardcode xterm…”
   • [community] “I have tried every possible setting but SSH ends up breaking more often than not. As opposed to iTerm which just works. Sometimes I'm back t…”
   • [community] “Missing search and weird ssh control character issues are my blockers. It's great otherwise!”

   proof ↗⚑ flag

Tmux

1. developerUse tmux inside the terminal, or a documented native tmux integration/control mode

   weight 2 · round drawn
   Warp—none0/10

   The evidence pack contains no mention of tmux support, tmux control mode, or any documented native tmux integration in Warp's docs or community discussion — Warp's session/pane features (blocks, session restoration) are its own proprietary system, not tmux. Since terminal emulators are a fair category for this axis, absence of evidence yields 'none'.

   ⚑ flag

   Ghostty—none0/10

   The evidence pack contains no mention of tmux compatibility, tmux control mode, or any native tmux integration in Ghostty's docs or community reports; only SSH-related terminfo/control-character friction is discussed, which is a different issue. Since terminal emulators like iTerm2 do document explicit tmux control-mode support, this is a fair axis to expect evidence for, and none exists here.

   • [community] “Using its own TERM is a deliberate design decision that can cause SSH 'missing or unsuitable terminal' errors on servers that hardcode xterm…”
   • [community] “I have tried every possible setting but SSH ends up breaking more often than not. As opposed to iTerm which just works. Sometimes I'm back t…”
   • [community] “Missing search and weird ssh control character issues are my blockers. It's great otherwise!”

   proof ↗⚑ flag

Agent control

1. ai-native userLet a coding agent programmatically drive the terminal itself — create panes, run commands, read output — through a documented control protocol or API

   weight 3 · round to Warp
   Warp~partialprobed6/10

   Warp exposes a genuine programmatic surface — an OpenAPI spec (confirmed live at docs.warp.dev/openapi.json), an Oz CLI with API keys for headless/CI control, MCP as a plugin interface, and a documented Warp Agent CLI that spins up its own PTY so an agent can run commands, provide input, and control full-screen apps (warp-docs-69, warp-docs-24, warp-docs-32). This supports agent-driven terminal automation, but the evidence never documents a protocol/API specifically for creating or managing Warp's terminal panes/windows or reading raw pane output — the API/CLI focus is on running and orchestrating agents rather than exposing the terminal UI itself as a controllable surface. Missing for 10: documented pane-creation/window-management API, explicit read-output API distinct from agent conversation transcripts, and independent/hands-on confirmation of external programmatic terminal control.

   • [claimed-docs] “API, CLI, SDK and MCP built as a platform, not a vertical product or AI teammate.”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”
   • [claimed-docs] “It creates and maintains its own pseudoterminal (PTY), which lets you and the agent execute commands and interact with running processes thr…”
   • [claimed-docs] “You can run full-screen and interactive terminal apps like sqlite and python through the agent session. Warp Agent can also control those ap…”
   • [probe] “PROBE openapi: HTTP 200 at https://docs.warp.dev/openapi.json — contains "openapi" key”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”

   proof ↗⚑ flag

   Ghostty~partialprobed4/10

   Ghostty documents a built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events, which could let an agent create panes and send commands on macOS, but this is a niche, platform-limited mechanism, not a documented cross-platform control API, and there's no evidence it supports reading terminal output. A hands-on runtime probe explicitly found no remote-control/IPC interface beyond local CLI keybinding actions, meaning an external agent cannot send commands to or read output from a running instance via any documented protocol. missing for 10: cross-platform control API, documented output-reading mechanism, and confirmation that AppleScript scripting actually supports agent-driven read/write loops.

   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”

   proof ↗⚑ flag

Extensibility

1. developerExtend the terminal with scripts or plugins in a real language (Lua, Python, kittens) beyond built-in options

   weight 2 · round drawn
   Warp—none0/10

   Warp's extensibility is centered on MCP servers (for extending AI agent tool access) and a settings file, not a general-purpose scripting/plugin language like Lua or Python for customizing terminal behavior (e.g., kitty-style 'kittens'). No evidence pack item documents a scripting API, plugin SDK, or embedded language runtime for extending the terminal itself.

   • [claimed-docs] “MCP servers extend Warp’s local agents in a modular, flexible way by exposing custom tools or data sources through a standardized interface …”
   • [claimed-docs] “we are shipping a (long-overdue) settings file so that users and agents get programmatic control over settings and easy portability between …”

   proof ↗⚑ flag

   Ghostty—none0/10

   Ghostty offers only configuration-file options, keybindings, and macOS-only AppleScript automation for driving the app externally; there is no evidence of a Lua/Python scripting API, plugin system, or kitten-style extension mechanism. The runtime probe explicitly confirms no remote-control/IPC surface exists beyond a fixed set of bindable actions, meaning scripts/plugins in a real language cannot extend the terminal.

   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [claimed-docs] “Ghostty supports flexible, custom keybindings through the keybind configuration option.”

   proof ↗⚑ flag

Layouts

1. power-userDefine startup sessions and window layouts in config or scripts so a project workspace opens in one command

   weight 2 · round to Ghostty
   Warp—none0/10

   Warp documents automatic session restoration (windows/tabs/panes reopened on relaunch) but this is a global auto-restore behavior, not a user-authored config/script that defines named per-project startup layouts launchable via one command. No evidence of a project-workspace/layout definition file or CLI flag to open a specific named session.

   • [claimed-docs] “Restore your windows, tabs, panes, and recent Blocks automatically when you relaunch Warp.”
   • [claimed-docs] “Session restoration allows you to quickly pick up where you left off in your previous terminal session.”

   proof ↗⚑ flag

   Ghostty~partialprobed3/10

   Ghostty offers rich config options, keybindings, and multi-window/tab/split support, plus a macOS-only AppleScript dictionary for 'scripting windows, tabs, terminals, layouts' which could approximate scripted startup layouts, but there is no documented session/layout file format or CLI flag to launch a predefined multi-window/tab/split workspace in one command (and probe evidence shows no external IPC/remote-control interface). missing for 10: cross-platform session/layout config file, documented 'launch project workspace' example, evidence of one-command startup replicating saved layouts, and confirmation the AppleScript route actually achieves this in practice.

   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Ghostty supports multiple windows, each with its own tabs and splits. These are all rendered using native UI components.”
   • [claimed-docs] “Windows, tabs, and splits: Ghostty supports multiple windows, each with its own tabs and splits.”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”

   proof ↗⚑ flag

Remote control

1. developerScript and control the terminal from the command line — open windows, send text, query state — via a documented CLI or IPC interface

   weight 3 · round to Warp
   Warp~partialclaimed5/10

   Warp ships documented CLI/IPC surfaces — the standalone Warp Agent CLI (works in any terminal, over SSH, creates its own PTY, can run/control full-screen apps) and the Oz CLI for running/managing cloud agents from scripts or CI with API-key auth — giving real command-line scriptability of agent sessions. However, the evidence is agent-centric (starting conversations, running commands, controlling processes) rather than direct terminal-window primitives like opening/arranging windows or querying pane/window state as the story literally describes (tmux/iTerm2-style control). Missing for 10: explicit documented commands to open/manage Warp windows or panes, send raw text to a specific pane, or query terminal UI state independent of an agent session; independent hands-on confirmation of such scripting.

   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline.”
   • [claimed-docs] “The Oz CLI is the command-line tool for running and managing Warp’s cloud agents from any terminal, script, or CI pipeline. Use it to start …”
   • [claimed-docs] “With the Oz CLI, you can: Run agents locally for development and debugging, Run agents on remote machines, Connect agents to MCP servers lik…”
   • [claimed-docs] “API Keys - Create and manage API keys to authenticate the Oz CLI without human interaction, ideal for CI pipelines, headless servers, and co…”
   • [claimed-docs] “Because it’s a regular command-line program, it works in any terminal emulator, including over SSH.”
   • [claimed-docs] “It creates and maintains its own pseudoterminal (PTY), which lets you and the agent execute commands and interact with running processes thr…”
   • [claimed-docs] “The Warp Agent CLI lets you switch directories while in an agent session, and even run agents on remote machines without installing remote b…”
   • [claimed-docs] “The Warp Agent is now available as a standalone CLI that you can use anywhere – Ghostty, iTerm 2, VS Code, built-in Windows and Mac Terminal…”

   proof ↗⚑ flag

   Ghostty~partialprobed4/10

   Ghostty documents an AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events (macOS only), and exposes CLI flags mapping to config keys, but hands-on probing found no cross-platform CLI/IPC mechanism to remote-control a running instance — the CLI's +action surface only lists/binds actions locally, it cannot send text or query state of a live terminal from another process. missing for 10: a documented cross-platform IPC/socket/CLI remote-control protocol, evidence of sending text/querying state from an external script, and Linux/Windows equivalent to AppleScript automation.

   • [claimed-docs] “AppleScript Automation: Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Built-in AppleScript dictionary for scripting windows, tabs, terminals, layouts, and input events.”
   • [claimed-docs] “Every configuration key is a valid CLI flag when launching Ghostty from the command-line.”
   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”

   proof ↗⚑ flag

Command tracking

1. developerI get command-aware features from shell integration, like exit status, duration, and notifications when long commands finish

   weight 2 · round to Warp
   Warp~partialcommunity4/10

   Only evidence is a community anecdote naming notifications on long-running command completion as a favorite feature; there is no first-party documentation confirming exit status display, command duration tracking, or notification configuration as shell-integration features (the docs pack is dominated by AI/agent capabilities instead). missing for 10: official docs on exit status indicators, duration display, and notification settings; independent corroboration beyond a single forum comment.

   • [community] “My daily driver for a few months now, even after trying Rio/Ghostty/Wezterm... Favorite feature is the notification when a long running comm…”

   proof ↗⚑ flag

   Ghostty—none0/10

   The evidence pack contains no mention of shell integration, exit status tracking, command duration, or completion notifications for Ghostty; only general feature docs (themes, keybindings, quick terminal, etc.) are present.

   ⚑ flag

Marks

1. developerJump between prompts and command marks in my scrollback thanks to shell integration

   weight 2 · round drawn
   Warp—none0/10

   The evidence pack references Warp's 'Blocks' concept (grouping commands and output) and session restoration of blocks, but never documents a shell-integration feature for jumping/navigating between prompt marks or command blocks in scrollback via keyboard shortcuts. Missing for 10: explicit documentation of prompt/command mark navigation shortcuts, scrollback jump-to-previous/next-prompt functionality, and any shell-integration-specific navigation UI.

   • [claimed-docs] “Restore your windows, tabs, panes, and recent Blocks automatically when you relaunch Warp.”
   • [claimed-docs] “Session restoration allows you to quickly pick up where you left off in your previous terminal session.”
   • [claimed-docs] “Sometimes, you may want to prevent a sensitive Block from being saved on your computer, or you may want to clear blocks from a machine entir…”

   proof ↗⚑ flag

   Ghostty—none0/10

   No evidence in the pack mentions shell integration, prompt/command marks, or scrollback navigation features (e.g., jump-to-prompt) for Ghostty; community feedback even highlights basic scrollback/search gaps without mentioning mark navigation.

   ⚑ flag

Discoverability

1. power-userDiscover and run terminal actions from a searchable command palette

   weight 1 · round to Warp
   Warp~partialclaimed5/10

   Warp Drive lets users save workflows and run them 'on-demand' (warp-docs-11, warp-docs-30), and AI Command Suggestions via '#' provide natural-language command discovery (warp-docs-8), which together approximate a command-palette-style workflow, but no evidence explicitly describes a searchable palette UI for discovering and running arbitrary terminal actions. Missing for 10: explicit documentation of a dedicated command palette (e.g., Cmd+P style fuzzy search across commands/workflows), independent/community confirmation of this specific feature, and clarity on scope (built-in commands vs only saved workflows).

   • [claimed-docs] “Warp lets you save your most important workflows in Warp Drive so you can run them on‑demand or share with your team.”
   • [claimed-docs] “Warp lets you save your most important workflows in Warp Drive so you can run them on‑demand or share with your team. Let Warp AI handle the…”
   • [claimed-docs] “Type '#' on your command line and start describing the command you want to run using natural language. Warp will load AI Command Suggestions…”

   proof ↗⚑ flag

   Ghostty—none0/10

   Evidence shows Ghostty exposes keybindings, a config-driven action system, and a CLI `+list-actions`/`+help` surface, but nothing describes a searchable/discoverable command palette UI for running terminal actions interactively; the runtime probe explicitly notes there's no remote-control/interactive command surface beyond static CLI listing.

   • [probe] “PROBE runtime (recorded 2026-09-04, Ghostty 1.3.1 via brew cask): `ghostty +list-actions` enumerates 85 bindable actions (new_split, new_tab…”
   • [claimed-docs] “Ghostty supports flexible, custom keybindings through the keybind configuration option.”

   proof ↗⚑ flag

Hyperlinks

1. developerOpen URLs and file paths from output by clicking them or selecting them from the keyboard

   weight 2 · round drawn
   Warp—none0/10

   The evidence pack contains no documentation or community mention of clickable URLs/file paths, keyboard-selectable links, or similar output-navigation features in Warp's terminal Blocks. This is a fair ergonomics question for a terminal product but nothing in the pack confirms it ships.

   ⚑ flag

   Ghostty—none0/10

   The evidence pack contains no mention of clickable URL/path detection in terminal output; the closest feature (proxy icon drag) only concerns the current session's working directory in the title bar, not clicking links or paths within terminal output. Missing for 10: any documentation or community mention of URL/path hyperlink detection, click-to-open behavior, or keyboard-based link selection.

   ⚑ flag

Scrollback

1. developerSearch my scrollback quickly and jump between matches

   weight 2 · round drawn
   Warp—none0/10

   The evidence pack covers Warp's AI/agent features, session restoration, theming, and telemetry, but contains no mention of scrollback search or match navigation functionality, despite this being a reasonable and expected axis for any terminal emulator.

   ⚑ flag

   Ghostty—none0/10

   No documentation feature for scrollback search exists, and multiple independent user reports confirm cmd-f/ctrl-f search does not work in Ghostty, calling it a blocker.

   • [community] “Ghostty is awesome and I almost dropped iTerm for it until I hit cmd-f and nothing happened.”
   • [community] “Missing search and weird ssh control character issues are my blockers. It's great otherwise!”
   • [community] “Missing scrollbars. That and ctrl-f are my two annoyances, apart from that I love ghostty and use it daily at work.”

   proof ↗⚑ flag