agentleFS
ddivyanshsharmaaddivyanshsharmaa/fable-toolkit/skills/team/SKILL.md

team

Skill0 starsChanged 44 days agoMITView on GitHub

Hand a task to the AI team with one request. Uses the engines the user said they have (Claude Code, the Codex CLI, the Antigravity CLI), checks they are signed in, then runs the right loop and assigns every part automatically: Claude plans and verifies, Codex takes precise coding and code or security audits, Antigravity takes bulk generation, web research, and content or SEO audits. Never asks who should do what. Use when the user says "use the team", "ai team", "the team", "delegate this", "get Codex or Antigravity on this", "full team", "reconfigure the team", or hands over a big job with several parts.

50 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use community-led-growth, ph-community-outreach and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • nicepkgSkill

    community-led-growth

    Expert in community-led growth (CLG) - leveraging user communities to drive acquisition, retention, and expansion. Covers building developer communities, user groups, ambassador programs, and turning customers into advocates. Knows the difference between community as a feature vs community as a growth engine, and how to measure community ROI. Use when "community-led, CLG, ambassador program, champion program, user community, developer community, word of mouth, user group, " mentioned.

    Also in 5 projects that use this

  • nicepkgSkill

    ph-community-outreach

    Engage Reddit, Indie Hackers, Hacker News and other communities for Product Hunt launches. Use this skill to identify relevant communities, craft appropriate posts, and drive cross-platform engagement without appearing spammy.

    Also in 5 projects that use this

  • xfstudioSkill

    micro-saas-launcher

    Expert in launching small, focused SaaS products fast - the indie hacker approach to building profitable software. Covers idea validation, MVP development, pricing, launch strategies, and growing to sustainable revenue. Ship in weeks, not months. Use when: micro saas, indie hacker, small saas, side project, saas mvp.

    Also in 5 projects that use this

  • mastepanoskiSkill

    ux-audit-rethink

    Comprehensive UX audit using IxDF's 7 factors, 5 usability characteristics, and 5 interaction dimensions. Holistic evaluation with redesign proposals based on user-centered design principles.

    Also in 5 projects that use this

  • ibahgatSkill

    ralplan

    Create a strict, quality-first execution plan with explicit gates and rollback-ready sequencing.

    Also in 8 projects that use this

  • ncklrsSkill

    seo-content-strategist

    Expert SEO content strategy for SaaS and B2B sites. Use when doing keyword research, planning content clusters, creating pillar pages, optimizing on-page SEO, fixing technical SEO issues, building links, refreshing content, implementing programmatic SEO, mapping search intent, optimizing for SERP features, or measuring content performance. Use for SEO audits, content gap analysis, and organic growth strategy.

    Also in 5 projects that use this

  • +6 more from nicepkg

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 3 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 36 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the bundled check through the Bash tool:

    line 25
  2. Very specific:

    If there is no Bash tool (native Windows without Git for Windows, where Claude Code runs PowerShell instead), the Codex and Antigravity …

    line 44
  3. Very specific:

    If FABLE_CONFIGURED is no (the user installed through /plugin and never ran the installer), ask once, with the detected state as the …

    line 56
  4. Very specific:

    After that, never ask again.

    line 65
  5. Distinctive:

    Never block the task on an optional fix.

    line 85
  6. Distinctive:

    Keep for yourself anything that touches credentials, production data, or an irreversible action.

    line 129

“Common” is measured: kept by at least 6 in 10 of the 36 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/team, one request, the right engines, automatically

The user should never need to know which engine does what. In this skill you find out what is available, fix or flag anything in the way, pick the loop, assign every unit to the engine that fits it, and prove the assembled result works.

Small job? If you could finish it in a handful of your own tool calls, just do it and skip the team. The team pays off when work splits into parts, runs long, or needs a second model's precision.

Step 1. Preflight, once per session

Run the bundled check through the Bash tool:

bash "${CLAUDE_SKILL_DIR}/scripts/preflight.sh"

It only reads, never changes anything, and takes a few seconds (up to a minute when Antigravity is slow to list its models). The report ends with machine readable lines:

LineValues
FABLE_MODEtrio, duo-codex, duo-gemini, or solo
FABLE_CODEX, FABLE_AGYready, not-signed-in, unreachable, missing, or off (the user's saved setup says not to use it)
FABLE_AGY_FASTAntigravity model id for bulk and mechanical work
FABLE_AGY_HEAVYmodel id for long generation, where drift and truncation are the risk
FABLE_AGY_SMARTmodel id for hard logic, security review, and escalation
FABLE_CONFIGUREDyes when the user has saved which engines they use, no otherwise
FABLE_FIXEShow many setup fixes the report recommends

If there is no Bash tool (native Windows without Git for Windows, where Claude Code runs PowerShell instead), the Codex and Antigravity loops cannot run: both CLIs wait forever unless stdin is closed with </dev/null, which PowerShell 5.1 cannot do. Run the job in solo mode with orchestrate, and tell the user once: install Git for Windows (https://git-scm.com/downloads/win) and restart Claude Code, and their saved engines take over from then on. Do not change the saved setup.

The saved setup is the user's decision, and you never change it on your own. It lives in fable-toolkit.conf in the Claude Code config folder, written by the installer or by configure.sh. An engine set to no is never used, even when installed. An engine set to yes that is not ready this time (signed out, offline) is skipped for this run only; say so in one line with the fix, and leave the file alone so the engine comes back by itself once it is ready.

If FABLE_CONFIGURED is no (the user installed through /plugin and never ran the installer), ask once, with the detected state as the suggested answer: "Do you use Codex?" and "Do you use Antigravity?". Use AskUserQuestion when it is available. Save the answers and rerun the preflight:

bash "${CLAUDE_SKILL_DIR}/scripts/configure.sh" --codex <yes|no> --agy <yes|no>
bash "${CLAUDE_SKILL_DIR}/scripts/preflight.sh"

After that, never ask again. The one exception is the user asking to change it ("reconfigure the team", "I have Codex now", "stop using Antigravity"): ask the two questions again, or take the answer straight from their message, and run configure.sh with it.

Then tell the user the mode in one line, for example: "Running as trio: Codex and Antigravity are both signed in." Do not ask which engines they have beyond the one time above.

Step 2. Apply the setup fixes the user will want

The report lists two kinds of fix.

  • Sign-ins (codex login, or running agy once). These open a browser, so only the user can finish them. Give the exact command, say which mode you are running in meanwhile, and carry on in that mode rather than waiting.
  • Claude Code settings (a permission rule so dispatches stop prompting; on macOS, Linux, and WSL2 with the sandbox on, excluding the two CLIs from it). Show the exact JSON change and ask once. On yes, apply it yourself: copy settings.json to a timestamped backup next to it, merge the keys without removing anything already there, confirm the file still parses as JSON, and tell the user /status shows whether it loaded.

Never block the task on an optional fix. The loops work without them; the fixes only remove friction.

Step 3. Route to the loop

FABLE_MODEFollow this skillWorkers
triotrioCodex and Antigravity
duo-codexduo-codexCodex
duo-geminiduo-geminiAntigravity
soloorchestrateClaude subagents

Invoke that skill (in a plugin install its name carries a fable-toolkit: prefix), or read its instructions directly at ${CLAUDE_SKILL_DIR}/../<skill>/SKILL.md. Follow it from its Setup step, skip its own preflight since you already ran it, and use the model ids from this preflight wherever it names an Antigravity model.

Step 4. Assign every unit automatically

Decide per unit from the signal in the work. Never ask the user who should do it; say who did what in the final report.

Signal in the unittrioduo-codexduo-geminisolo
Subtle logic, edge cases decide correctnessCodexCodexAntigravity smartClaude worker
Must read existing code and match its patternsCodexCodexAntigravity smartClaude worker
Code review or security audit of a repositoryCodex, --sandbox read-onlyCodex, --sandbox read-onlyAntigravity smart, then your own passClaude critic
Same mechanical change across many filesAntigravity fastCodex, split into parallel unitsAntigravity fastClaude workers
Long generated docs, data, config, or contentAntigravity heavyCodexAntigravity heavyClaude worker
Web research, competitor or market scan, SEO or content audit of public pagesAntigravityClaude researcherAntigravityClaude researcher
Reading or summarising many local filesAntigravity, files named in the briefCodex, read-onlyAntigravityClaude worker
A product decision, a taste call, or anything without a checkable criterionyouyouyouyou

"Claude worker", "Claude researcher", and "Claude critic" are this toolkit's worker, researcher, and critic agents, spawned with the Agent tool.

Antigravity can search the web and fetch pages in headless mode, so research and public-page audits go to it without any setup. It cannot run shell commands there, so anything that needs a command (a crawl, a test run, a git log) you run first and hand over as files.

For research and audit units, the brief must demand a source (URL, or file and line) for every finding, and your review checks at least three of those sources yourself. A finding without a source does not count.

Keep for yourself anything that touches credentials, production data, or an irreversible action. Briefs go to external services.

Step 5. Finish the way the loop says

The chosen loop's review, fix rounds, final gate, and report all apply unchanged. Add one line to the report naming which engine handled which units, so the user sees the split without asking.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.