Agent skill

Eng Worktree Isolation

by compozy in compozy/compozy

Isolates Compozy runtime state for concurrent worktrees and QA runs.

MITAuto-check passedDevelopment

Install Eng Worktree Isolation

skills CLI
$ npx skills add compozy/compozy --skill eng-worktree-isolation -a claude-code

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

GitHub CLI
$ gh skill install compozy/compozy eng-worktree-isolation --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/compozy/compozy.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/eng/eng-worktree-isolation .claude/skills/eng-worktree-isolation && 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
eng-worktree-isolation
GitHub stars
2.8k
Token cost
~1.5k tokens
SKILL.md length
751 words
Files
2 (incl. scripts)
Skills in repo
47
Repo updated
First seen
Licence
MIT

At a glance

Isolates Compozy runtime state for concurrent worktrees and QA runs.

  • Works in 3 steps: Look for explicit signals in the user's… → If no concurrency signal is present and… → Confirm the scenario-slug, defaulting to…
  • Runtime may share the machine
  • SKILL.md covers Required Inputs, Procedures and Error Handling
  • Runs Python scripts from its folder; calls make, python3 and bun

What it does

Eng Worktree Isolation is an agent skill from compozy/compozy. Isolates Compozy runtime state for concurrent worktrees and QA runs. Use when another agent or runtime may share the machine, or when creating a parallel worktree. Do not use for single-worktree or build-only work.

Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including scripts (for example `scripts/allocate-isolation.py`).

It sits in Development, covering Git worktrees. The repository describes itself as: An operating system for AI agents. Plug in the agent CLIs you already use (Claude Code, Codex, Gemini CLI, Cursor) and they become a team: they split the work, hand tasks to each… The licence is MIT.

When your agent uses it

  • Runtime may share the machine
  • Creating a parallel worktree
  • Single-worktree
  • Build-only work

Example prompts

  • “Use the eng-worktree-isolation skill to isolate Compozy runtime state for concurrent worktrees and QA runs”
  • “/eng-worktree-isolation”

Requirements

  • Python 3

Workflow steps

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

  1. Look for explicit signals in the user's request: parallel-worktree language, "another agent is running", "tem outro agent trabalhando"…
  2. If no concurrency signal is present and no parallel runtime is planned, stop; this skill does not apply.
  3. Confirm the scenario-slug, defaulting to a timestamped slug when omitted.

What it can do on your machine

Read from SKILL.md and the folder at commit c46be43. 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

    Ships 1 file in scripts/ (Python), which the agent can run.

    Shell commands in SKILL.md call:

    • make
    • python3
    • bun

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

  • Network

    No URLs in SKILL.md.

    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

Eng Worktree Isolation loads about 1.5k tokens when it runs. Until then it costs about 59 tokens; SKILL.md has 751 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~59
When it runs · the whole SKILL.md, loaded when a task matches
~1.5k

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); the scripts in this folder are not scanned.

SKILL.md

The full file from compozy/compozy at commit c46be43, republished under its MIT licence (© compozy). 751 words, ~1,518 tokens.

Download SKILL.mdSave it as .claude/skills/eng-worktree-isolation/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
eng-worktree-isolation
description
Isolates Compozy runtime state for concurrent worktrees and QA runs. Use when another agent or runtime may share the machine, or when creating a parallel worktree. Do not use for single-worktree or build-only work.
trigger
explicit
argument-hint
[scenario-slug]

Worktree Isolation

Provision one owned runtime envelope for a concurrent worktree or non-bootstrap test run. Production-like QA should prefer eng-qa-bootstrap, which adds a manifest, provider homes, browser policy, and targeted teardown evidence.

Creating the parallel checkout itself is a separate step: make worktree-new SLUG=<slug> (→ ./scripts/worktree.sh) adds a sibling worktree and bootstraps it (mise pins, bun install + skill/AGENTS.md symlinks; BUILD=1/E2E=1 opt-ins); make worktree-bootstrap preps an existing checkout; ./scripts/worktree.sh rm <slug> removes. Go caches are shared machine-wide — no per-worktree Go setup. This skill then isolates the RUNTIME state that agents inside that worktree use. Opt-in make gate-full and local E2E self-queue machine-wide (L-030); required scoped lanes need no lock and PR CI owns full verification.

Required Inputs

  • scenario-slug (optional): a short kebab-case slug used to name the COMPOZY_HOME directory and tmux socket. Defaults to compozy-iso-<timestamp>.

Procedures

Step 1: Confirm the Concurrency Branch

  1. Look for explicit signals in the user's request: parallel-worktree language, "another agent is running", "tem outro agent trabalhando", "QA in parallel", or invocation under a worktree path like Compozy/_worktrees/<slug>/.
  2. If no concurrency signal is present and no parallel runtime is planned, stop; this skill does not apply.
  3. Confirm the scenario-slug, defaulting to a timestamped slug when omitted.

Done when: the concurrent runtime/test branch and one unique scenario slug are explicit.

Step 2: Allocate COMPOZY_HOME

  1. Run the bootstrap/mutating helper python3 .agents/skills/eng/eng-worktree-isolation/scripts/allocate-isolation.py --slug "<scenario-slug>" [--prefer-worktree]. Pass --prefer-worktree only from an actual parallel worktree. The script:
    • Creates a unique COMPOZY_HOME directory under ${TMPDIR:-/tmp}/compozy-iso-<slug>-<random>/ OR uses the worktree-scoped Compozy/_worktrees/<slug>/.compozy/ when invoked from a worktree.
    • Picks a free TCP port on 127.0.0.1 for the daemon HTTP server.
    • Creates a unique UDS path under COMPOZY_HOME.
    • Picks a unique tmux socket path under the COMPOZY_HOME (e.g., ${COMPOZY_HOME}/tmux-bridge.sock).
  2. The script prints export statements, including COMPOZY_ISOLATION_ROOT, suitable for eval "$(...)".

Done when: the allocator returns one non-default, owned root and unique HTTP/UDS/tmux addresses.

Step 3: Source the Envelope

  1. Capture the Step 2 allocator output once and source those exported variables: COMPOZY_ISOLATION_ROOT, COMPOZY_HOME, COMPOZY_HTTP_PORT, COMPOZY_UDS_PATH, TMUX_BRIDGE_SOCKET.
  2. Shells may combine allocation and sourcing in one invocation (do not rerun this after Step 2 already allocated an envelope): eval "$(python3 .agents/skills/eng/eng-worktree-isolation/scripts/allocate-isolation.py --slug "<slug>" [--prefer-worktree])".
  3. For Make/CI invocations: pass the variables as overrides to the daemon start command.
  4. Confirm the daemon does NOT write to ~/.compozy/ or default port 23230.

Done when: the current shell exposes the exact allocator output and no default runtime path is reachable by the planned command.

Step 4: Verify Isolation Before Action

  1. Confirm COMPOZY_HOME is non-default and writable.
  2. Confirm the chosen ports are not already bound (re-pick if necessary).
  3. Confirm the tmux socket path is non-default and not held by another process.
  4. Print a one-line summary: slug, COMPOZY_HOME, http port, uds path, tmux socket.
Show full SKILL.md (293 more words)Show less

Done when: every address is free immediately before launch and the summary identifies the owned root.

Step 5: Run the Isolated Scenario

  1. Hand off to the inner skill (eng-real-scenario-qa, qa-execution, make test-e2e-runtime, etc.).
  2. Prefer eng-qa-bootstrap for production-like local QA because its own allocator also provides provider homes, a manifest, browser policy, Web proxy env, and teardown evidence.
  3. Inner skills inherit the env via the shell session. Do not re-allocate.

Done when: every child process uses the same envelope and none writes to another worktree or default home.

Step 6: Cleanup (processes ALWAYS, files optionally)

  1. Process teardown is mandatory on every terminal path. Run the mutating teardown helper only for this run's envelope: python3 .agents/skills/eng/eng-qa-bootstrap/scripts/teardown-qa-env.py --root "$COMPOZY_ISOLATION_ROOT"
  2. Confirm the targeted teardown reports TEARDOWN_ALL_CLEAN=true. Survivors are a blocking failure.
  3. The COMPOZY_HOME directory is left in place for forensic inspection unless the caller explicitly requests --purge for a temporary envelope.
  4. Never purge a worktree-scoped .compozy; it belongs to the user's checkout.
  5. Use make qa-reap or teardown --all only for an intentional machine-wide stale-lab recovery, never as normal cleanup while other runs may be active.

Done when: the owned envelope is clean, unrelated labs remain untouched, and any retained files contain no live process state.

Error Handling

  • No free port available: retry with a wider range. If still no luck, surface the busy ports and exit.
  • COMPOZY_HOME path collision: the script uses random suffixes; collision is essentially impossible. If it happens, retry once.
  • User invokes without concurrency signal but with --force: apply isolation. Some users always want isolated runs.
  • Worktree-scoped path lacks write permission: fall back to TMPDIR-scoped path with a logged warning.
  • Targeted teardown cannot prove ownership: stop and report the root/PIDs; never widen normal cleanup to --all.

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

Files

SKILL.md and 1 other file (scripts) in .agents/skills/eng/eng-worktree-isolation of compozy/compozy.

  • SKILL.md
  • scripts/allocate-isolation.py

Open the folder on GitHubat commit c46be43

Compare with similar skills

Eng Worktree Isolation 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.

Eng Worktree Isolation compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Eng Worktree Isolation this skillcompozy/compozy2.8k—~1.5kAutomated safety check: PassMIT
Finishing a Development Branchobra/superpowers296k5 repos~1.9kAutomated safety check: PassMIT
Migrate Core Code to Submodulestinyhumansai/openhuman41k—~2.6kAutomated safety check: PassGPL-3.0
Finishing A Development Branchfarm-fe/farm5.6k33 repos~1.8kAutomated safety check: PassMIT
Git Worktree Cleanuplobehub/lobehub83k—~2.8kAutomated safety check: PassCustom licence
Keep Codex Fastvibeforge1111/keep-codex-fast1.6k—~3.1kAutomated safety check: PassMIT

Similar skills

  • Walks the last step of a branch: confirm tests pass, detect the git environment, ask how to integrate, carry out your choice and clean up the worktree.

    296k GitHub starsUsed in 5 repos~1.9k tokens
    DevelopmentAuto-check passed
  • Migrate Core Code to Submodules

    tinyhumansai/openhuman

    Plans and carries out moving non-host-specific code and its tests from the OpenHuman core into vendored tiny submodule libraries, then releases the submodule and re-pins the host.

    41k GitHub stars~2.6k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • A skill your agent uses when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for…

    5.6k GitHub starsUsed in 33 repos~1.8k tokens
    DevelopmentAuto-check passed
  • Git Worktree Cleanup

    lobehub/lobehub

    Audits stale Git worktrees and branches with a bundled script, classifies each one, and deletes only after you approve the exact candidates.

    83k GitHub stars~2.8k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Keep Codex Fast

    vibeforge1111/keep-codex-fast

    A skill your agent uses when Codex feels slow or bloated, when local sessions/logs/worktrees/config have grown over time, or when a user wants safe maintenance for Codex Desktop/CLI state.

    1.6k GitHub stars~3.1k tokensUpdated 5 mo ago
    DevelopmentAuto-check passed
  • Pre-Release PR Triage

    jamiepine/voicebox

    Sorts a backlog of open pull requests into must-merge, candidate, superseded and deferred, writes a triage doc and works the merge loop before a release.

    57k GitHub stars~3.1k tokensUpdated yesterday
    DevelopmentAuto-check passed

More from compozy/compozy

All 47 skills in this repo
  • Eng Real Scenario QA

    compozy/compozy

    Dogfoods Compozy through an autonomous startup scenario with live providers, cross-surface observation, and strict evidence audit.

    2.8k GitHub stars~3.4k tokensUpdated today
    Auto-check passed
  • Eng Test Conventions

    compozy/compozy

    Go test-shape discipline for Compozy. An agent skill from compozy/compozy.

    2.8k GitHub stars~467 tokensUpdated today
    Auto-check passed
  • Assistant UI

    compozy/compozy

    Guide for assistant-ui library - AI chat UI components. An agent skill from compozy/compozy.

    2.8k GitHub stars~958 tokensUpdated today
    Auto-check passed
  • Guide for assistant-ui UI primitives - ThreadPrimitive, ComposerPrimitive, MessagePrimitive.

    2.8k GitHub stars~999 tokensUpdated today
    Auto-check passed
  • Assistant UI Runtime

    compozy/compozy

    Guide for assistant-ui runtime system and state management. An agent skill from compozy/compozy.

    2.8k GitHub stars~856 tokensUpdated today
    Auto-check passed
  • Assistant UI Streaming

    compozy/compozy

    Guide for assistant-stream package and streaming protocols. An agent skill from compozy/compozy.

    2.8k GitHub stars~813 tokensUpdated today
    Auto-check passed

Categories

Questions about Eng Worktree Isolation

What does Eng Worktree Isolation do?

Isolates Compozy runtime state for concurrent worktrees and QA runs. Eng Worktree Isolation is an agent skill from compozy/compozy. Isolates Compozy runtime state for concurrent worktrees and QA runs.

When should I use Eng Worktree Isolation?

Eng Worktree Isolation fits situations like: runtime may share the machine; creating a parallel worktree; single-worktree; build-only work.

How do I install Eng Worktree Isolation in Claude Code?

Run `npx skills add compozy/compozy --skill eng-worktree-isolation -a claude-code`. Or copy the skill folder (.agents/skills/eng/eng-worktree-isolation in compozy/compozy) into .claude/skills/eng-worktree-isolation in your project. Claude Code loads it when a task matches its description.

How do I install Eng Worktree Isolation in Codex?

Run `npx skills add compozy/compozy --skill eng-worktree-isolation -a codex`. Or copy the skill folder (.agents/skills/eng/eng-worktree-isolation in compozy/compozy) into .agents/skills/eng-worktree-isolation in your project. Codex loads it when a task matches its description.

Can I use Eng Worktree Isolation 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 compozy/compozy --skill eng-worktree-isolation -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/eng-worktree-isolation, .gemini/skills/eng-worktree-isolation, .github/skills/eng-worktree-isolation and .opencode/skills/eng-worktree-isolation in your project.

What does Eng Worktree Isolation need to run?

Going by SKILL.md and its folder, Eng Worktree Isolation needs Python for the scripts in its folder and the command-line tools its instructions call (make, python3 and bun). Our summary lists: Python 3.

Does Eng Worktree Isolation access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Eng Worktree Isolation 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. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Eng Worktree Isolation use?

Eng Worktree Isolation 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 Eng Worktree Isolation use?

About 1.5k tokens (SKILL.md is roughly 6.1k 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 Eng Worktree Isolation?

Skills that share tags, products or a category with Eng Worktree Isolation: Finishing a Development Branch (obra/superpowers, 296k stars), Migrate Core Code to Submodules (tinyhumansai/openhuman, 41k stars), Finishing A Development Branch (farm-fe/farm, 5.6k stars) and Git Worktree Cleanup (lobehub/lobehub, 83k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Eng Worktree Isolation?

compozy (a GitHub organization) maintains it in compozy/compozy, which has 2,791 GitHub stars. The repository holds 47 skills in this directory. The repository was last updated on October 7, 2026.

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