turbocharge
CLAUDE.md2 starsChanged 11 months agoMITView on GitHub
Turbocharge: Claude Code Plugin Blueprint A Claude Code plugin that replaces ad-hoc agents, scattered skills, and custom commands with a single opinionated engineering pipeline. Also serves as a reference implementation for building Claude Code plugins. This is a plugin for Claude Code — not a standalone app. It consists of: - 10 skills (SKILL.md files that define slash commands) - 5 agents (markdown agent definitions dispatched by skills) - 1 hook (SessionStart bootstrap + ATLAS pre-load) - A marketplace manifest…
We split this CLAUDE.md into the rules it gives an agent
5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 83
Every skill MUST have:
- Distinctive: line 91
Sonnet minimum, never Haiku: Every agent (builders, reviewers, researchers, planners) runs Sonnet or higher.
- Distinctive: line 115
Use semver: patch for fixes, minor for new skills/features, major for breaking changes
- Very specific: line 132
Update CHANGELOG.md with the new entry
- Very specific: line 133
Commit + push to nicodiansk/turbocharge.
≈ “Common” is an estimate: at least 1 in 5 of the 25,680 CLAUDE.md 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.
Files
CLAUDE.md
Project Context
Turbocharge: Claude Code Plugin Blueprint
A Claude Code plugin that replaces ad-hoc agents, scattered skills, and custom commands with a single opinionated engineering pipeline. Also serves as a reference implementation for building Claude Code plugins.
What This Project Is
This is a plugin for Claude Code — not a standalone app. It consists of:
- 10 skills (SKILL.md files that define slash commands)
- 5 agents (markdown agent definitions dispatched by skills)
- 1 hook (SessionStart bootstrap + ATLAS pre-load)
- A marketplace manifest for distribution
Architecture
Plugin Structure:
skills/ → SKILL.md files (user-invocable slash commands)
agents/ → Agent definitions (dispatched by skills, not invocable directly)
hooks/ → hooks.json + content files (lifecycle automation)
.claude-plugin/ → plugin.json manifest
settings.json → Plugin-scoped settingsThe Pipeline
brainstorm → story → plan → build → review → ship
| |
debug wrapDistribution
Source repo: nicodiansk/turbocharge (this repo — single source of truth)
Install: claude plugin marketplace add nicodiansk/turbocharge
claude plugin install turbocharge@turbocharge
Update: claude plugin update turbocharge@turbocharge
Local dev: claude --plugin-dir /path/to/turbocharge
Directory: claude.ai plugin directory listing "Turbocharge" (ours; submitted 2026-06-23,
published 2026-09-24). Managed at claude.ai/directory/manage; tracks `master`.
Each new version is scanned, then needs "Publish update" (a reviewer makes it live).Name history: 3.0.0–3.1.1 shipped as turboflow because we wrongly believed another author owned
turbocharge in the directory; the listing was ours. 3.2.0 restored turbocharge. Never start a new
directory submission. Never rename the plugin name: the listing and every install key off it.
Domain Terms
| Term | Definition |
| Skill | A SKILL.md file in skills/<name>/ — becomes /turbocharge:<name> |
| Agent | A markdown file in agents/ — dispatched by skills via the Agent tool |
| Hook | Shell command executed at lifecycle events (SessionStart) |
| Marketplace | GitHub repo with .claude-plugin/marketplace.json that indexes plugins — lives at the root of this repo (nicodiansk/turbocharge) as of v2.3.0; previously a sibling turbocharge-marketplace repo (deleted 2026-04-13) |
| Lane | A group of plan tasks that owns a disjoint set of files (plan's ## Lanes table). /turbocharge:build runs each lane inline in the main session or with one Sonnet builder, overlaps lanes with no After dependency, commits by pathspec, and runs the full suite once at the join |
| Iron Law | An enforced constraint (not a suggestion) baked into skill definitions |
| Red Flag | Anti-rationalization table that catches Claude skipping process steps |
| ATLAS.md | Semantic domain map generated by /turbocharge:atlas in user projects (Where to Look, Entry Points, Module Map, Key Symbols, Integration Points, Conventions & Gotchas). Where to Look table pre-loaded by SessionStart hook (remaining sections available via Read on demand); injected into dispatch prompts for planner, researcher, and code-reviewer agents; nudged by SessionStart when absent; staleness detected via <!-- atlas-hash:XXXX --> footer. This plugin repo intentionally does NOT ship one — CLAUDE.md + README cover the same ground for a plugin blueprint. |
Development Conventions
File Types
| File | Format | Purpose |
skills/<name>/SKILL.md | Markdown with YAML frontmatter | Skill definition (prompt template) |
agents/<name>.md | Plain markdown | Agent system prompt |
hooks/hooks.json | JSON | Hook registration (lifecycle → command mapping) |
hooks/*.md | Markdown | Hook content (injected into context) |
.claude-plugin/plugin.json | JSON | Plugin manifest (name, version, author) |
settings.json | JSON | Plugin-scoped Claude Code settings |
Skill Anatomy
Every skill MUST have:
- YAML frontmatter with
name,description - Clear instructions Claude can follow without ambiguity
- A "Red Flags" section for anti-rationalization (build, review, debug at minimum)
- Chain-forward: suggest the next skill in the pipeline
Agent Models
Sonnet minimum, never Haiku. Every agent (builders, reviewers, researchers,
planners) runs Sonnet or higher. Turn count beats token price: a cheaper model
that loops on misunderstandings costs more than one that gets it right the first
time. Do not set model: haiku on any agent.
The planner is the one deliberate model: inherit: it rides the orchestrator's
session model so architectural planning gets your strongest available model
(Opus when you orchestrate on Opus, Sonnet otherwise) — never below the Sonnet
floor in normal use, and portable to setups without Opus access. Every other
agent hard-pins model: sonnet.
Release Naming
Name releases by what they do, not after the source of inspiration. Never put
competitor/plugin codenames (e.g. another plugin's brand) in commit messages,
branch names, CHANGELOG entries, or PR titles/bodies. Continue the project's own
lineage instead (e.g. "Lean Builder v3" → "Lean review pipeline"). The one allowed
exception is the pre-existing setup skill's conflict-detection list, which names
competing plugins functionally to warn about conflicts.
Versioning
- Version lives only in
.claude-plugin/plugin.json;marketplace.jsoncarries none (plugin.json wins over the marketplace entry, and the docs say to set it in one place). Guarded byt_version_lockstep.sh - Bump it every release — the claude.ai directory expects a new version for each update
- Use semver: patch for fixes, minor for new skills/features, major for breaking changes
Testing Locally
# Load plugin from local directory
claude --plugin-dir .
# Validate plugin structure
./scripts/validate.shPublishing Flow
Single-repo as of v2.3.0. No cross-repo sync.
- Bump
versionin.claude-plugin/plugin.json(the only place it lives) - Update
CHANGELOG.mdwith the new entry - Commit + push to
nicodiansk/turbocharge. Users pick it up viaclaude plugin update turbocharge@turbocharge.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.

