New: 97,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Delegate Verbose

by timusus in timusus/Shuttle2

Run high-volume work in a subagent so its raw output never enters this conversation.

Apache-2.0Auto-check passedMobile

Install Delegate Verbose

skills CLI
$ npx skills add timusus/Shuttle2 --skill delegate-verbose -a claude-code

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

GitHub CLI
$ gh skill install timusus/Shuttle2 delegate-verbose --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/timusus/Shuttle2.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/delegate-verbose .claude/skills/delegate-verbose && 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
delegate-verbose
GitHub stars
229
Token cost
~864 tokens
SKILL.md length
484 words
Files
1
Skills in repo
13
Repo updated
First seen
Licence
Apache-2.0

At a glance

Run high-volume work in a subagent so its raw output never enters this conversation.

  • Works in 2 steps: The output will exceed roughly 2k… → You only need the conclusion — not the…
  • Run instrumented tests
  • SKILL.md covers When to delegate, When NOT to delegate, How and Pick the model for the subagent, plus 1 more section
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Delegate Verbose is an agent skill from timusus/Shuttle2. Run high-volume work in a subagent so its raw output never enters this conversation. Use BEFORE running a Gradle test sweep, instrumented/emulator test run, or lint sweep — anything where the output is large but only the conclusion matters. Triggers on "run the tests", "gradle test", "run instrumented tests", "run lint", "run all tests".

Its SKILL.md is about 860 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Mobile, covering Subagents and Linting and formatting. It works with Gradle and Android. The repository describes itself as: Shuttle Music: a music player for Android and iOS. Plays local files and streams from Jellyfin, Emby, Plex and Navidrome/Subsonic, with Chromecast, Android Auto, EQ and replay… The licence is Apache-2.0.

When your agent uses it

  • Run instrumented tests
  • Tasks that involve Subagents
  • Tasks that involve Linting and formatting

Example prompts

  • “run the tests”
  • “gradle test”
  • “run instrumented tests”
  • “/delegate-verbose”

Workflow steps

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

  1. The output will exceed roughly 2k tokens, and
  2. You only need the conclusion — not the raw text — to continue.

What it can do on your machine

Read from SKILL.md and the folder at commit 116c579. 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

    No scripts in the folder and no shell commands in SKILL.md.

    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

Delegate Verbose loads about 864 tokens when it runs. Until then it costs about 89 tokens; SKILL.md has 484 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~89
When it runs · the whole SKILL.md, loaded when a task matches
~864

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 timusus/Shuttle2 at commit 116c579, republished under its Apache-2.0 licence (© timusus). 484 words, ~864 tokens.

Download SKILL.mdSave it as .claude/skills/delegate-verbose/SKILL.md (or your agent's skills folder).
name
delegate-verbose
description
Run high-volume work in a subagent so its raw output never enters this conversation. Use BEFORE running a Gradle test sweep, instrumented/emulator test run, or lint sweep — anything where the output is large but only the conclusion matters. Triggers on "run the tests", "gradle test", "run instrumented tests", "run lint", "run all tests".

Delegate verbose work

Raw tool output stays in context for the rest of the session and is resent on every subsequent turn. A 40k-token test log read once costs 40k tokens on every turn that follows. Delegating it costs 40k once, inside the subagent, and returns a summary.

When to delegate

Delegate when both are true:

  1. The output will exceed roughly 2k tokens, and
  2. You only need the conclusion — not the raw text — to continue.

Concretely in this project: ./gradlew testDebugUnitTest sweeps across modules, support/scripts/emu-verify.sh --suite (emulator run), support/scripts/lint across the whole tree, and reading anything under build/reports/. support/scripts/unit-test already condenses its Gradle output through build-brief when it's on PATH (falling back to plain ./gradlew); a raw ./gradlew invocation still benefits from delegation since it has no condenser of its own.

When NOT to delegate

  • You need the exact bytes (a stack trace you'll edit against, a diff you'll apply).
  • The work is a single short command whose output you'd read in full anyway.
  • The task requires the conversation's accumulated context to interpret — a subagent starts cold and cannot see this thread.

How

Launch a subagent with the Agent tool. Give it the full command to run and, critically, tell it what to return — a subagent's final message is its return value, so a vague prompt yields a vague summary and you'll have to run it again.

Bad: "Run the tests and tell me how it went." Good: "Run support/scripts/unit-test. Return a list of every failing test class/method with its assertion message, which module it's in, and the overall pass/fail count. Do not paste raw log lines."

Ask for a bounded shape (a table, a ranked list, N bullet points), not "a summary".

If the subagent's answer turns out to be insufficient, send it a follow-up message rather than re-running the sweep in this context — it still holds the raw output.

Show full SKILL.md (171 more words)Show less

Pick the model for the subagent

A subagent inherits the parent's model unless told otherwise, so a delegated grep runs on Opus by default. Pass model explicitly:

  • haiku — mechanical and extractive. Run a command and report pass/fail, grep for a pattern, count occurrences, pull the failing lines out of a log.
  • sonnet — needs to read code but not to make a design judgement. Trace where a value comes from, summarise what a module does, draft a mechanical refactor.
  • omit it (inherit) — anything where being wrong is expensive: reviewing a diff for correctness, diagnosing a race, deciding whether an approach is sound.

The failure mode is asymmetric: too cheap means a re-run, which costs more than the model saving ever did. When genuinely unsure, omit it.

Verifying it worked

After the subagent returns, the raw output should be absent from this conversation. If you find yourself pasting its logs back in to reason about them, the delegation failed — the right fix is a sharper return contract next time, not abandoning the pattern.

© timusus, Apache-2.0. 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 .claude/skills/delegate-verbose of timusus/Shuttle2.

Open the folder on GitHubat commit 116c579

Compare with similar skills

Delegate Verbose 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.

Delegate Verbose compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Delegate Verbose this skilltimusus/Shuttle2229—~864Automated safety check: PassApache-2.0
Configuring Test Dependenciesskydoves/android-testing-skills333—~4.1kAutomated safety check: PassApache-2.0
Android CheckAdyen/adyen-android149—~357Automated safety check: PassMIT
Android API Diffgkd-kit/gkd43k—~796Automated safety check: PassGPL-3.0
Android Developmentdpconde/claude-android-skill336—~1.7kAutomated safety check: PassMIT
Run Jetpack Android Appwordpress-mobile/WordPress-Android3.2k—~886Automated safety check: PassGPL-2.0

Similar skills

  • Configuring Test Dependencies

    skydoves/android-testing-skills

    A skill your agent uses to wire the correct Gradle dependency matrix for Jetpack Compose UI tests.

    333 GitHub stars~4.1k tokensUpdated 4 mo ago
    MobileAuto-check passed

  • Android Check

    Adyen/adyen-android

    Run compile, lint, and unit test checks. An agent skill from Adyen/adyen-android.

    149 GitHub stars~357 tokensUpdated yesterday
    Testing & QAAuto-check passed

  • Android API Diff

    gkd-kit/gkd

    Looks up Android framework Java and AIDL APIs across versions with the android-api-diff CLI: signatures, availability, source files and hidden-API access code.

    43k GitHub stars~796 tokensUpdated yesterday
    MobileAuto-check passed

  • Android Development

    dpconde/claude-android-skill

    Create production-quality Android applications following Google's official architecture guidance and NowInAndroid best practices.

    336 GitHub stars~1.7k tokensUpdated 10 mo ago
    MobileAuto-check passed

  • Run Jetpack Android App

    wordpress-mobile/WordPress-Android

    Builds the Jetpack debug app with Gradle and installs it on a connected Android device or an emulator started from an available AVD.

    3.2k GitHub stars~886 tokensUpdated yesterday
    MobileAuto-check passed

  • Orca Android Emulator Control

    stablyai/orca

    Android device and emulator control from inside Orca over adb, with the live device view in Orca's emulator pane. Use when driving an adb-connected emulator…

    87k GitHub stars~558 tokensUpdated today
    MobileAuto-check passed

More from timusus/Shuttle2

All 13 skills in this repo

  • Android Coroutine Testing

    timusus/Shuttle2

    Comprehensive guidance for writing deterministic tests for Android code using Kotlin coroutines.

    229 GitHub stars~1.2k tokensUpdated yesterday
    Auto-check passed

  • Debug Receivers

    timusus/Shuttle2

    Drive the S2 debug build's playback and queue over ADB broadcasts — play the whole library, play/pause, skip, seek, remove a queue item, toggle shuffle/repeat, dump playback state as JSON, reimport…

    229 GitHub stars~1.9k tokensUpdated yesterday
    Auto-check passed

  • Deploy Android

    timusus/Shuttle2

    Deploy the Android app to the Play Store via GitHub Actions.

    229 GitHub stars~1.2k tokensUpdated yesterday
    Auto-check passed

  • Reply Reviews

    timusus/Shuttle2

    Weekly Play Store review flow for S2: fetch new reviews, draft replies, get owner approval, then post.

    229 GitHub stars~1k tokensUpdated yesterday
    Auto-check passed

  • Check

    timusus/Shuttle2

    Run lint + unit tests, investigate and fix failures. An agent skill from timusus/Shuttle2.

    229 GitHub stars~449 tokensUpdated yesterday
    Auto-check passed

  • Commit

    timusus/Shuttle2

    Analyse working tree changes, group them into logical atomic commits, and create well-structured conventional commits with module scopes.

    229 GitHub stars~1.9k tokensUpdated yesterday
    Auto-check: notes

Works with

Categories

Questions about Delegate Verbose

What does Delegate Verbose do?

Run high-volume work in a subagent so its raw output never enters this conversation. Delegate Verbose is an agent skill from timusus/Shuttle2. Run high-volume work in a subagent so its raw output never enters this conversation.

When should I use Delegate Verbose?

Delegate Verbose fits situations like: run instrumented tests; tasks that involve Subagents; tasks that involve Linting and formatting.

How do I install Delegate Verbose in Claude Code?

Run `npx skills add timusus/Shuttle2 --skill delegate-verbose -a claude-code`. Or copy the skill folder (.claude/skills/delegate-verbose in timusus/Shuttle2) into .claude/skills/delegate-verbose in your project. Claude Code loads it when a task matches its description.

How do I install Delegate Verbose in Codex?

Run `npx skills add timusus/Shuttle2 --skill delegate-verbose -a codex`. Or copy the skill folder (.claude/skills/delegate-verbose in timusus/Shuttle2) into .agents/skills/delegate-verbose in your project. Codex loads it when a task matches its description.

Can I use Delegate Verbose 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 timusus/Shuttle2 --skill delegate-verbose -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/delegate-verbose, .gemini/skills/delegate-verbose, .github/skills/delegate-verbose and .opencode/skills/delegate-verbose in your project.

What does Delegate Verbose need to run?

SKILL.md names no scripts, command-line tools or credentials: Delegate Verbose is instructions for the agent only.

Does Delegate Verbose 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 Delegate Verbose 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 Delegate Verbose use?

Delegate Verbose is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Delegate Verbose use?

About 864 tokens (SKILL.md is roughly 3.5k 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 Delegate Verbose?

Skills that share tags, products or a category with Delegate Verbose: Configuring Test Dependencies (skydoves/android-testing-skills, 333 stars), Android Check (Adyen/adyen-android, 149 stars), Android API Diff (gkd-kit/gkd, 43k stars) and Android Development (dpconde/claude-android-skill, 336 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Delegate Verbose?

timusus (a GitHub user) maintains it in timusus/Shuttle2, which has 229 GitHub stars. The repository holds 13 skills in this directory. The repository was last updated on October 6, 2026.

Source: timusus/Shuttle2 on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.