Agent skill

Codewhale Dogfood Install

by codewhale-hq in codewhale-hq/Codewhale

Proves a Codewhale change in the real product: a stamped release build, an atomic local install, fresh-shell verification and manual QA that automated gates cannot cover.

MITAuto-check passedDevelopment

Install Codewhale Dogfood Install

skills CLI
$ npx skills add codewhale-hq/Codewhale --skill cw-dogfood -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install codewhale-hq/Codewhale cw-dogfood --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/codewhale-hq/Codewhale.git skills-src && mkdir -p .claude/skills && cp -r skills-src/docs/skills/cw-dogfood .claude/skills/cw-dogfood && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
cw-dogfood
GitHub stars
41k
Token cost
~1.3k tokens
SKILL.md length
628 words
Files
1
Skills in repo
63
Repo updated
First seen
Licence
MIT

At a glance

Proves a Codewhale change in the real product: a stamped release build, an atomic local install, fresh-shell verification and manual QA that automated gates cannot cover.

  • Works in 7 steps: Gate first. Run cw-gates to the rung the… → Build stamped. Local builds are… → Install atomically. Use the script; do… → …
  • Proving a user-visible TUI or approval-flow change in the running product
  • SKILL.md covers When to use, Workflow, Red flags / don't and Output
  • Calls git and cargo

What it does

The skill covers the step where a Codewhale change is proved in the real product rather than only by passing gates, which show that code compiles and asserts but miss freezes, focus theft, streaming cadence and approval-flow regressions. It puts the actual binary on your PATH and has the agent use it. It is stage four of a six-skill loop that runs orient, slice, gates, dogfood, land and handoff.

The workflow starts by running the gates, never installing an ungated build, then builds a release binary stamped with the current commit SHA, since the installer refuses unstamped builds. `scripts/release/install-dogfood.sh` installs `codewhale` and `codew` atomically and refuses a dirty source tree unless overridden, and the agent must never copy over a running binary because that can hang later launches on Apple Silicon. It then verifies from a fresh login shell that the version string contains the short HEAD SHA, and exercises the changed behavior in a real terminal.

When your agent uses it

  • Proving a user-visible TUI or approval-flow change in the running product
  • Installing a local build of Codewhale for manual testing
  • Before calling a runtime bug fixed or landing a release candidate

Example prompts

  • “Dogfood this change: build the stamped release and install it on my machine.”
  • “Install the local build and verify from a fresh shell that codew reports the right commit.”
  • “Before we land this, run a manual QA pass on the new approval flow in the real binary.”

Requirements

  • A Codewhale source checkout with a Rust toolchain (cargo)
  • A real terminal for manual testing

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. Gate first. Run cw-gates to the rung the change
  2. Build stamped. Local builds are unstamped ((dev)), and the
  3. Install atomically. Use the script; do not hand-copy
  4. Verify from a fresh shell, not this one. A correct target/release
  5. Use the product. Run it in a real terminal and exercise what you changed.
  6. Headless surfaces, when the change touches them. codewhale exec is the
  7. Record what you saw. Dimensions, inputs, visible state, side effects. A

What it can do on your machine

Read from SKILL.md and the folder at commit 0ea319a. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git
    • cargo

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Codewhale Dogfood Install loads about 1.3k tokens when it runs. Until then it costs about 60 tokens; SKILL.md has 628 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~60
When it runs · the whole SKILL.md, loaded when a task matches
~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from codewhale-hq/Codewhale at commit 0ea319a, republished under its MIT licence (© codewhale-hq). 628 words, ~1,322 tokens.

Download SKILL.mdSave it as .claude/skills/cw-dogfood/SKILL.md (or your agent's skills folder).
name
cw-dogfood
description
Use when a Codewhale change needs proving in the real product, or when asked to build/install/dogfood the local binaries: stamped release build, atomic install, fresh-shell verification, and the manual QA that gates cannot cover.

cw-dogfood

Green gates prove the code compiles and asserts. They do not prove the product works. Freezes, route contamination, focus theft, streaming cadence, and approval-flow regressions all live in the runtime, where no unit test looks. This stage puts the actual binary on your PATH and makes you use it.

Stage 4 of the loop: cw-orient → cw-slice → cw-gates → dogfood → cw-land → cw-handoff.

When to use

  • The change is user-visible: TUI layout, motion, streaming, model or Fleet selection, approvals, commands, install paths.
  • Before landing a release candidate, or before claiming a runtime behavior is fixed.
  • The user asks to "install the build", "dogfood this", or "get this on my machine".

Workflow

  1. Gate first. Run cw-gates to the rung the change deserves. Never install an ungated build.

  2. Build stamped. Local builds are unstamped ((dev)), and the installer refuses an unstamped binary on purpose — the stamp is what proves the thing on your PATH is the thing you just built:

    bash
    CODEWHALE_BUILD_SHA=$(git rev-parse HEAD) \
      cargo build --release --locked -p codewhale-cli --bin codewhale
  3. Install atomically. Use the script; do not hand-copy:

    bash
    scripts/release/install-dogfood.sh          # defaults to target/release

    It refuses a dirty source tree (override deliberately with CODEWHALE_ALLOW_DIRTY_DOGFOOD=1, and then say so wherever you report the install), verifies the binary embeds current HEAD, installs codewhale and codew into ~/.cargo/bin and ~/.local/bin (override with CODEWHALE_INSTALL_DIRS), re-signs ad-hoc on macOS, and verifies resolution from a fresh login shell.

    Never cp over a running binary. On Apple Silicon that poisons the kernel's code-signature cache for the inode, and later execs hang until reboot. The installer does tmp-copy plus atomic mv for exactly this reason.

  4. Verify from a fresh shell, not this one. A correct target/release binary and a stale codew on PATH is the classic false pass:

    bash
    zsh -lc 'type -a codew codewhale; codew --version'

    The version string must contain the short HEAD SHA you just built.

  5. Use the product. Run it in a real terminal and exercise what you changed. crates/tui/AGENTS.md owns sizes, motion evidence, and environment caveats; judge motion against docs/MOTION_CONTRACT.md.

    Scenarios worth exercising when they are in scope:

    • Liveness under fanout — spawn several workers; typing, render, cancel, and the roster stay live throughout, and Esc cancels mid-fanout.
    • Route isolation — multiple terminals on distinct provider/model routes, zero cross-terminal contamination, no provider+model mismatch.
    • Running-turn input — during a busy turn, Enter queues a follow-up, an empty Enter promotes the oldest, Ctrl+Enter steers, Shift+Enter newlines.
    • Approvals — ordinary tool approval vs. repository-law approval; the screen must name the repository constitution where it applies, and decorative motion must go still when the user owns the next action.
    • Empty, narrow, and first-run states — compact layouts remove chrome before content.
  6. Headless surfaces, when the change touches them. codewhale exec is the one-shot worker path; codewhale app-server is the local control/API surface. They must agree about routing, permissions, and event states — a disagreement is a runtime bug, not a QA note.

    bash
    scripts/release/app-server-smoke.sh
    codewhale exec --auto --output-format stream-json --model <model> "Reply PONG"

    Provider calls spend tokens. Ask before running the paid ones.

  7. Record what you saw. Dimensions, inputs, visible state, side effects. A screenshot proves layout and color; only live observation or a recording proves motion and continuity.

Show full SKILL.md (135 more words)Show less

Red flags / don't

  • Don't cp a binary over a running one. Use install-dogfood.sh.
  • Don't verify in the shell that already has the old binary resolved.
  • Don't claim animation or cadence quality from a still image.
  • Don't substitute a full-screen assertion harness for looking at and using the product.
  • Don't install from a dirty tree without saying so in the report — the version stamp will not show the dirt.
  • Don't delete target/dogfood/* bundles if the lane keeps them: they are release evidence.
  • Don't spend provider tokens on smoke runs without approval.

Output

  • The exact build command, including the CODEWHALE_BUILD_SHA stamp.
  • The installer's destinations and its fresh-shell verification result.
  • codew --version from a fresh login shell, with the SHA visible.
  • Per scenario: terminal size, what you did, what you observed — and which scenarios you did not exercise.

© codewhale-hq, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in docs/skills/cw-dogfood of codewhale-hq/Codewhale.

Open the folder on GitHubat commit 0ea319a

Compare with similar skills

Codewhale Dogfood Install next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Codewhale Dogfood Install compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Codewhale Dogfood Install this skillcodewhale-hq/Codewhale41k—~1.3kAutomated safety check: PassMIT
TUI Change Verification for Warpwarpdotdev/warp65k1 repos~5.5kAutomated safety check: NotesAGPL-3.0
Pre-Merge Checkjsmastery-pro/skills1.5k—~1.1kAutomated safety check: NotesMIT
Fix What I Pointed Atreticlehq/reticle1.2k—~878Automated safety check: PassApache-2.0
OpenLogi macOS Permissions TriageAprilNEA/OpenLogi23k—~2.5kAutomated safety check: NotesApache-2.0
Apple Container Test RunnerRustPython/RustPython22k—~467Automated safety check: PassMIT

Similar skills

  • Runs Warp's terminal UI for real, drives it with keystrokes and reads the rendered screen back to confirm that a change looks and behaves as intended.

    65k GitHub starsUsed in 1 repo~5.5k tokens
    Testing & QAAuto-check: notes
  • Pre-Merge Check

    jsmastery-pro/skills

    A gate before merge: verify runs the real app against the spec, and review has a different model do a senior code review, without editing code.

    1.5k GitHub stars~1.1k tokensUpdated 2 mo ago
    DevelopmentAuto-check: notes
  • Fix What I Pointed At

    reticlehq/reticle

    Picks up bugs a person flagged by pointing at elements in the running app, each mark carrying the element, their note and the source file and line, then fixes and verifies them.

    1.2k GitHub stars~878 tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Decides whether an OpenLogi device problem on macOS is a privacy-permission (TCC) problem, using agent log lines, and says which identity needs which grant.

    23k GitHub stars~2.5k tokensUpdated today
    DevelopmentAuto-check: notes
  • Apple Container Test Runner

    RustPython/RustPython

    Runs RustPython tests inside a Linux container built with Apple's container CLI, so macOS users can compare Linux results with their local ones.

    22k GitHub stars~467 tokensUpdated today
    Testing & QAAuto-check passed
  • PlotJuggler Live Verification

    PlotJuggler/PlotJuggler

    Confirms that a change to PlotJuggler 4 really works in the running app by proving the rebuild, launching with real data and measuring the result.

    6.2k GitHub stars~956 tokensUpdated 9 days ago
    Testing & QAAuto-check passed

More from codewhale-hq/Codewhale

All 63 skills in this repo
  • Codewhale Session Handoff

    codewhale-hq/Codewhale

    Writes a paste-ready handoff for the next agent session, opening with a state-check command block and separating done, suspected and blocked work.

    41k GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Codewhale Landing Workflow

    codewhale-hq/Codewhale

    Decides how verified work should reach main, directly, in a worktree or on an integration branch, while keeping contributor credit and respecting merge gates.

    41k GitHub stars~1.6k tokensUpdated today
    Auto-check passed
  • Sub-Agent Delegation

    codewhale-hq/Codewhale

    Guides when and how to split multi-step coding, research or verification work into focused sub-agent runs while the parent keeps integration and final checks.

    41k GitHub stars~790 tokensUpdated today
    Auto-check passed
  • Codewhale Fleet Manager

    codewhale-hq/Codewhale

    Triages and manages Codewhale fleet runs and workers with typed commands, classifying failures and choosing a safe restart, resume or escalation.

    41k GitHub stars~1.2k tokensUpdated today
    Auto-check passed
  • GitHub Issue Bulk Assigner

    codewhale-hq/Codewhale

    Moves a list of GitHub issues into a milestone or assigns them to owners with the gh CLI, checking each one before and after the change.

    41k GitHub stars~953 tokensUpdated today
    Auto-check passed
  • Codewhale Issue Filing

    codewhale-hq/Codewhale

    Turns a bug or idea into a single actionable Codewhale GitHub issue with evidence, a gap-naming title, acceptance criteria and related links.

    41k GitHub stars~1.3k tokensUpdated today
    Auto-check passed

Works with

Questions about Codewhale Dogfood Install

What does Codewhale Dogfood Install do?

Proves a Codewhale change in the real product: a stamped release build, an atomic local install, fresh-shell verification and manual QA that automated gates cannot cover. The skill covers the step where a Codewhale change is proved in the real product rather than only by passing gates, which show that code compiles and asserts but miss freezes, focus theft, streaming cadence and approval-flow regressions. It puts the actual binary on your PATH and has the agent use it.

When should I use Codewhale Dogfood Install?

Codewhale Dogfood Install fits situations like: proving a user-visible TUI or approval-flow change in the running product; installing a local build of Codewhale for manual testing; before calling a runtime bug fixed or landing a release candidate.

How do I install Codewhale Dogfood Install in Claude Code?

Run `npx skills add codewhale-hq/Codewhale --skill cw-dogfood -a claude-code`. Or copy the skill folder (docs/skills/cw-dogfood in codewhale-hq/Codewhale) into .claude/skills/cw-dogfood in your project. Claude Code loads it when a task matches its description.

How do I install Codewhale Dogfood Install in Codex?

Run `npx skills add codewhale-hq/Codewhale --skill cw-dogfood -a codex`. Or copy the skill folder (docs/skills/cw-dogfood in codewhale-hq/Codewhale) into .agents/skills/cw-dogfood in your project. Codex loads it when a task matches its description.

Can I use Codewhale Dogfood Install in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add codewhale-hq/Codewhale --skill cw-dogfood -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cw-dogfood, .gemini/skills/cw-dogfood, .github/skills/cw-dogfood and .opencode/skills/cw-dogfood in your project.

What does Codewhale Dogfood Install need to run?

Going by SKILL.md and its folder, Codewhale Dogfood Install needs the command-line tools its instructions call (git and cargo). Our summary lists: A Codewhale source checkout with a Rust toolchain (cargo); A real terminal for manual testing.

Does Codewhale Dogfood Install access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Codewhale Dogfood Install safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Codewhale Dogfood Install use?

Codewhale Dogfood Install is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Codewhale Dogfood Install use?

About 1.3k tokens (SKILL.md is roughly 5.3k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Codewhale Dogfood Install?

Skills that share tags, products or a category with Codewhale Dogfood Install: TUI Change Verification for Warp (warpdotdev/warp, 65k stars), Pre-Merge Check (jsmastery-pro/skills, 1.5k stars), Fix What I Pointed At (reticlehq/reticle, 1.2k stars) and OpenLogi macOS Permissions Triage (AprilNEA/OpenLogi, 23k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Codewhale Dogfood Install?

codewhale-hq (a GitHub organization) maintains it in codewhale-hq/Codewhale, which has 41,082 GitHub stars. The repository holds 63 skills in this directory. The repository was last updated on October 11, 2026.

Source: codewhale-hq/Codewhale on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.