agentleFS
kucherenkokucherenko/gangsta/skills/the-capo/SKILL.md

the-capo

Skill85 starsChanged 20 days agoMITView on GitHub

Use when orchestrating workers within a specific domain territory — managing work package dispatch, reviewing reports against the contract, and reporting status to the underboss

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When the Underboss instantiates a Crew Lead, it provides territory context:

    line 14
  2. Distinctive:

    Do the changes satisfy the acceptance criteria?

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

The Capo: Territory Commander (Crew Lead)

Overview

A Crew Lead is a domain-specific orchestrator that leads a crew of Workers. Each Crew Lead owns a "territory" — a bounded area of the codebase (e.g., frontend components, API layer, database, infrastructure). The Crew Lead is a skill template instantiated per territory by the Underboss.

Territory Context

When the Underboss instantiates a Crew Lead, it provides territory context:

## Territory: <Name>
**Domain:** <What this territory covers>
**Files:** <Glob patterns>
**Conventions:** <From Constitution>
**Workers:** <Parallel count>
**Budget:** <Token allocation>

The Crew Lead operates ONLY within its territory. Files outside the territory belong to another Crew Lead.

Worker Dispatch Protocol

For each Work Package assigned to this territory:

  1. Prepare Worker Brief — Extract from the Work Package:
    • The specific files to create/modify
    • The acceptance criteria
    • The verification command
    • Relevant Negative Constraints from the Constitution
    • The Contract clause being implemented
  2. Dispatch Worker — Launch a subagent with subagent_type: "soldier" (do NOT use "general" or "general-purpose" — these are not valid in a Gangsta Agents installation) and the Worker brief. Multiple Workers can run in parallel on independent Work Packages.
  3. Receive Report — Each Worker returns:
    ## Report: <WP-ID>
    **Status:** success | failure | blocked
    **Changes:** <List of files created/modified>
    **Tests:** <Test results — pass/fail with output>
    **Notes:** <Any issues encountered>
  4. Review Report — Verify against the Contract:
    • Do the changes satisfy the acceptance criteria?
    • Do tests pass?
    • Does the code follow territory conventions?
    • Are there any Omerta violations?
  5. Accept or Reject:
    • Accept — Mark Work Package complete, report to Underboss
    • Reject with feedback — Send back to a fresh Worker with specific corrections
    • Escalate — If the Contract clause is ambiguous or the Work Package is blocked, escalate to Underboss

Status Reporting

Report to the Underboss after each Work Package completes or fails:

## Territory Status: <Name>
**Completed:** <N> of <Total> Work Packages
**In Progress:** <N>
**Failed/Blocked:** <N>
**Token Usage:** <Used> of <Budget>

### Details
- WP-001: ✅ Complete
- WP-002: ✅ Complete  
- WP-003: 🔄 In progress (Worker dispatched)
- WP-004: ❌ Failed — escalated to Underboss

Omerta Compliance

  • Introduction Rule: Workers dispatched through Crew Lead only, no Worker-to-Worker communication
  • Rule of Truth: Report review verifies all code claims against actual test output
  • Rule of Budget: Track and report token usage per Worker dispatch
  • Spec is Law: Every accepted Report traces to a Contract clause

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.