Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/skullninja/coco-workflow/claude-mdgit clone --depth 1 https://github.com/skullninja/coco-workflowWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/skullninja/coco-workflow/claude-md)<a href="https://agentmods.dev/instructions/skullninja/coco-workflow/claude-md"><img src="https://agentmods.dev/badge/instructions/skullninja/coco-workflow/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.07598 | $0.07598 |
| Opus 5 | $0.03799 | $0.03799 |
| Sonnet 5 | $0.01520 | $0.01520 |
| Haiku 4.5 | $0.00760 | $0.00760 |
Grade A, and why
coco-workflow CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 465 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md -- Coco
Project Overview
Coco is a Claude Code plugin that provides autonomous spec-driven development. It unifies planning (slash commands), execution (built-in task tracker), PR workflow with AI code review, and visibility (configurable issue tracker) into a single plugin distributed via the Claude Code marketplace (or as a git submodule).
Architecture
Five layers:
- Discovery:
/coco:prdand/coco:roadmapproduce PRD, analysis, and roadmap artifacts. Supports multi-repo via derived PRDs (/coco:prd derive) - Planning: Skills (
interview,design,tasks,import) produce spec artifacts inspecs/{feature}/ - Execution:
lib/tracker.sh(bash + jq) manages task state, dependencies, sessions - Review: Two-tier PR workflow with AI code review (
agents/code-reviewer.md) - Visibility: Issue tracker bridge (Linear MCP, GitHub CLI, or none) mirrors status
Key Files
| Path | Purpose |
|---|---|
.claude-plugin/plugin.json |
Claude Code plugin manifest (auto-discovers commands/skills/agents) |
hooks/hooks.json |
Claude Code hook definitions (PostToolUse, PreCompact, SessionStart) -- no PreToolUse, see Hooks |
bin/coco-tracker |
Tracker wrapper -- on PATH as bare command coco-tracker |
lib/tracker.sh |
Built-in task tracker -- core of the system (invoked via coco-tracker) |
config/coco.default.yaml |
Default configuration schema |
commands/setup.md |
/coco:setup -- project initialization (config wizard, git hooks, permissions) |
commands/ |
14 slash commands (setup, prd [greenfield/audit/derive], roadmap, phase, loop, execute, dashboard, standup, test-audit, etc.) |
commands/test-audit.md |
/coco:test-audit -- read-only test suite audit against the value rubric |
skills/interview/SKILL.md |
Pre-design discovery interview (AI-selected) |
skills/design/SKILL.md |
Feature design: spec + implementation plan (AI-selected) |
skills/tasks/SKILL.md |
Task list generation with consistency analysis (AI-selected) |
skills/import/SKILL.md |
Tracker + issue tracker import (AI-selected) |
skills/execute/SKILL.md |
Execution skill (delegates to /coco:execute command) |
skills/hotfix/SKILL.md |
Single-issue hotfix workflow (with optional PR) |
agents/code-reviewer.md |
AI code review agent for PRs |
agents/task-executor.md |
Worktree-isolated task executor for parallel execution |
agents/pre-commit-tester.md |
UI/UX validation agent (config-driven) |
agents/test-auditor.md |
Read-only test scoring agent, dispatched by /coco:test-audit for large suites |
hooks/scripts/post-tool-use-quality.sh |
PostToolUse hook -- runs lint/typecheck after file edits |
hooks/scripts/pre-compact.sh |
PreCompact hook -- captures session state before compaction |
hooks/scripts/session-start.sh |
SessionStart hook -- restores session context |
hooks/scripts/coco-lib.sh |
Shared hook helpers -- resolves the coco project root by walking up |
git-hooks/commit-msg.sh |
Commit message validation (reads config) |
git-hooks/pre-commit.sh |
Build check + UI change detection (reads config) |
GUIDE.md |
Comprehensive workflow guide with deep-dives and quick reference |
templates/ |
Default templates for PRD, analysis, roadmap, discovery, design, tasks, constitution |
templates/test-value-rubric.md |
Test value rubric -- verdicts and evidence used by review and audit |
scripts/setup.sh |
Creates .coco/ directory and installs git hooks in host project |
scripts/uninstall.sh |
Removes git hooks |
tests/test-tracker.sh |
53 tests for tracker.sh |
tests/test-hooks.sh |
9 tests for coco-lib.sh project-root resolution |
tests/test-setup-config.sh |
15 tests guarding wizard/config drift |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 465 lines · 7,598 tokens per session scan A 78e2f5c6c008
coco-workflow CLAUDE.md is an instructions file published in the GitHub repository skullninja/coco-workflow (7 stars, last pushed 4d ago), licensed MIT. It adds 7,598 tokens to every session, about $0.0380 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
agent-feed AGENTS.md
Instructions for fqmyysjjd/agent-feed, covering agent-feed ai development instructions, startup policy, rule priority, responsibility boundary and mandatory gates.
agent-feed CLAUDE.md
Instructions for fqmyysjjd/agent-feed, a project described as: Turn AI coding chats into a source-controlled workflow: AGENTS.md, rules, skills, verification gates, and adapters for Codex, Claude Code, and Cursor.
loop-lang CLAUDE.md
Claude Code instructions for tickets-forge-dev/loop-lang: This repo uses Loop — a small natural-language DSL for self-correcting, human-gated coding workflows. First time you touch Loop in this repo, read AGENTS.md end to end before authoring or running anything — it's the full, current grammar that ships with the…
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.