cc-rig 1.4.1

Project setup generator for Claude Code

Set up Claude Code the right way - in under a minute.

Getting Started · What Gets Generated · Stack & Process · For Teams · FAQ

---

A properly configured Claude Code project needs CLAUDE.md, hooks, agents, slash commands, skills, permissions, MCP servers and memory files. Each has specific formats, frontmatter and cross-references. That takes hours of manual work and most people skip it.

cc-rig generates all of it. Tell it what you're building and how you like to work. It writes 30-65 native Claude Code files tuned to your framework. No runtime dependency. No lock-in. Just files that Claude Code reads on startup.

---

Install

pip install cc-rig

• Python 3.9+. Zero external dependencies. Standard library only.
• Optional: pip install cc-rig[rich] for full-screen TUI wizard (arrow keys, radio buttons, checkboxes), colors, tables and progress bars. Auto-detected, falls back to stdlib ANSI if not installed.
• Works best with Claude Code v2.1.50+. Older versions or missing installs get a warning, but cc-rig generates everything anyway.

---

Getting Started

Start a new project

The interactive wizard walks you through it. With Textual installed, you get a full-screen TUI with arrow-key navigation, radio buttons and checkboxes. Without it, you get numbered CLI prompts.

cc-rig init

Or skip the wizard and specify everything directly:

cc-rig init --template fastapi --workflow standard --name my-api

Set up an existing project

Already have a codebase? cc-rig detects your stack from package.json, go.mod, Cargo.toml, pyproject.toml and more. It proposes what to add and won't touch existing files.

cd my-existing-project
cc-rig init --migrate

See it in action

Use a team config

A teammate already set up cc-rig? Load their config and get the same setup:

cc-rig init --config .cc-rig.json

Quick picker

Don't want the full wizard? Pick from numbered lists:

cc-rig init --quick

---

What Gets Generated

cc-rig generates native Claude Code files, the same formats from the official docs. Nothing proprietary. Delete cc-rig tomorrow and everything keeps working.

your-project/
├── CLAUDE.md                       # Project rules Claude follows
├── CLAUDE.local.md                 # Personal preferences (not git-tracked)
├── .mcp.json                       # MCP server integrations
├── .claude/
│   ├── settings.json               # Permissions, hooks, safety guards
│   ├── cc-rig-init.log             # Generation log (what was created + validation)
│   ├── agents/                     # Specialized agents for different tasks
│   ├── commands/                   # Slash commands you trigger with /
│   ├── hooks/                      # Auto-format, lint, safety blocks
│   └── skills/                     # Community skills auto-installed from 10 repos
├── agent_docs/                     # Framework-specific guides for Claude
└── memory/                         # Git-tracked team knowledge across sessions

Everything is tracked in a manifest, so cc-rig clean removes exactly what was generated. Nothing more.

CLAUDE.md

Targets under 100 lines. Static content first, dynamic content last. Claude Code's prompt cache is prefix-matched, so every wasted token costs money on every API call.

Includes project identity, stack, tool commands, guardrails, framework-specific rules and @import references to deeper docs (auto-loaded by Claude Code). Not a wall of text. A tight brief that Claude actually reads.

A companion CLAUDE.local.md is generated for personal preferences (not git-tracked). Use it for per-developer customization without affecting the shared config.

Agents

Isolated Claude instances in .claude/agents/, each with its own system prompt, model assignment and tool restrictions in YAML frontmatter. cc-rig emits up to 10 of Claude Code's 14 supported frontmatter fields — beyond the basics (name, description, model, tools), agents get optional fields like background, isolation, permissionMode, maxTurns and memory where appropriate. This means generated agents work out of the box with Claude Code's parallel execution features like /simplify and /batch.

Agent	Role	Model	Advanced fields
code-reviewer	6-aspect code review	Sonnet	memory: project
architect	System design, ADRs	Opus*	memory: project
test-writer	Test generation with coverage awareness	Sonnet	—
explorer	Fast codebase scanning	Haiku	permissionMode: plan, maxTurns: 15
refactorer	Safe refactoring with test verification	Sonnet	—
pr-reviewer	Pull request review	Opus*	—
security-auditor	OWASP-aware security review	Opus*	memory: project
implementer	Feature implementation from specs	Sonnet	—
doc-writer	Documentation generation	Sonnet	—
pm-spec	Specification creation from requirements	Opus*	—
techdebt-hunter	Technical debt identification	Sonnet	—
db-reader	Database schema and query analysis	Sonnet	—
parallel-worker	Background work in isolated git worktrees	Sonnet	background: true, isolation: worktree

*Opus on Max/Enterprise plans, Sonnet on Pro/Team. cc-rig auto-resolves model assignments based on your Claude plan tier — no manual config needed.

Your workflow preset determines which agents are included, from 3 for quick prototyping to 13 for full production rigor.

Slash Commands

Workflows you trigger with / in Claude Code. Each is a markdown file in .claude/commands/.

Command	What It Does
/fix-issue	Reproduce → diagnose → fix → test → commit
/plan	Architecture-first planning with checkpoints
/research	Explore codebase before implementing changes
/review	Multi-dimensional code review via agent
/test	Generate tests with coverage awareness
/assumptions	Surface Claude's hidden assumptions (with confidence levels)
/remember	Save learnings to persistent memory
/learn	Extract patterns from code for future sessions
/refactor	Safe refactoring with test verification
/optimize	Performance analysis and optimization
/techdebt	Identify and address technical debt
/spec-create	Create implementation spec from requirements
/spec-execute	Execute a spec with built-in validation
/daily-plan	Morning planning from active tasks
/worktree	Spawn a parallel worker in an isolated git worktree
/gtd-capture	Capture tasks into GTD inbox
/gtd-process	Process and prioritize captured tasks
/security	Security review via auditor agent
/document	Generate documentation

Up to 19 commands depending on your workflow preset.

Hooks

Shell scripts on Claude Code lifecycle events, configured in settings.json.

Event	What Fires	Why
PostToolUse (Write)	Auto-format (prettier/ruff/gofmt)	Instant cleanup, <1s
PreToolUse (Bash)	Lint + typecheck on git commit	Quality gate before commits
PreToolUse (Write/Bash)	Block rm -rf /, pushes to main, .env writes	Safety guards
Stop	Save learnings to memory, remind about tests, show session cost	Preserve context, cost awareness
PreCompact	Save context before compaction	Survive context loss
SessionStart	Load project context and active tasks	Continuity
SessionStart	Print open/done task counts (harness B1+)	Quick orientation
PreToolUse (Bash)	Lint gate on git commit (harness B2+)	Structural enforcement

Up to 14 hooks from your workflow preset, plus up to 3 more from the harness level.

Skills

Auto-invoked behaviors in .claude/skills/. Claude loads them when the task matches. No manual trigger needed.

The Claude Code skill ecosystem is huge — skills.sh indexes 73K+, repos like obra/superpowers, trailofbits/skills and anthropics/skills offer high-quality options across every SDLC phase. cc-rig gives you a smart starting set and makes it easy to add more.

Starter set (auto-installed at init):

• Framework-matched: Python projects get modern-python and property-based-testing, Next.js gets vercel-react-best-practices and next-best-practices, Go/Rust get static-analysis
• Cross-cutting: code review, security basics, TDD, debugging — scaled by workflow level (0 for speedrun, up to 14 for verify-heavy)
• project-patterns stub for your team's custom conventions

Optional skill packs (select during wizard or add later):

Pack	What it adds	Source repos
Security Deep Dive	supply chain auditing, variant analysis, dangerous API detection	trailofbits/skills
DevOps & IaC	Terraform, Kubernetes, monitoring, GitOps	hashicorp, ahmedasmar
Web Quality	Core Web Vitals, accessibility, SEO, performance	addyosmani
Code Quality	20 quality dimensions, anti-gaming scoring, scan→plan→fix loop	peteromallet/desloppify
Database Pro	migration patterns, query optimization, multi-DB support	multiple

Add skills anytime — from cc-rig or any source:

cc-rig skills list                    # Show what's installed
cc-rig skills catalog                 # Browse available packs and skills
cc-rig skills add <name>              # Install from catalog
cc-rig skills remove <name>           # Remove a skill
npx skills add <repo> --skill <name>  # Install any skill from any repo

Browse the full ecosystem: skills.sh · awesome-claude-skills · skillsmp.com

Plugins

cc-rig also curates 24 official Anthropic marketplace plugins and writes them into settings.json as enabledPlugins. Plugins are self-contained extensions that Claude Code installs and manages natively — no manual MCP setup or binary downloads.

Five categories:

Category	Count	Examples
LSP	7	pyright-lsp, typescript-lsp, gopls-lsp, rust-analyzer-lsp, jdtls-lsp, csharp-lsp, php-lsp
Integration	10	github, vercel, supabase, sentry, slack, linear, notion, firebase, gitlab, atlassian
Workflow	5	commit-commands, code-review, pr-review-toolkit, feature-dev, security-guidance
Autonomy	1	ralph-loop (official Anthropic autonomous iteration loop)
Utility	1	hookify (visual hook builder)

Smart defaults resolve plugins by language (LSP), template (integrations) and workflow (workflow plugins). The github plugin replaces the GitHub MCP server — self-contained, auto-start, no token config needed. Expert mode offers a multi-select from the full catalog.

Memory

Two complementary memory systems:

• Auto-memory (~/.claude/projects/) — personal, per-machine notes managed automatically by Claude Code. Always on. Use for personal preferences and local context.
• Team memory (memory/) — git-tracked shared knowledge that travels with the repo. Use for decisions, patterns, gotchas and conventions that every contributor should know.

cc-rig generates the team memory layer:

File	Purpose
decisions.md	Architectural decisions with rationale
patterns.md	Discovered code patterns and conventions
gotchas.md	Known issues, things that didn't work
people.md	Team ownership and responsibilities
session-log.md	Brief per-session progress log

A Stop hook prompts Claude to save team-relevant learnings before ending. A PreCompact hook does the same before context compaction wipes working memory. The /remember command routes personal notes to auto-memory and team knowledge to memory/ files.

Team memory files are not baked into CLAUDE.md. They load via Read tool on demand. This keeps the cached prompt prefix stable across sessions.

Permissions & Safety

settings.json with sensible defaults:

{
  "permissions": {
    "allow": [
      "Read", "Glob", "Grep", "Edit", "Write",
      "NotebookEdit", "Bash", "WebSearch", "WebFetch", "Task"
    ],
    "deny": [
      "Bash(rm -rf /)",
      "Bash(rm -rf ~)"
    ]
  }
}

Safety hooks handle the rest by blocking .env edits, pushes to main and destructive rm commands with exit 2.

MCP Servers

.mcp.json at the project root, configured for your stack. Available integrations:

Auto-configured (selected by template): PostgreSQL · Playwright

Available (add via expert mode): Slack · Linear · Sentry · Filesystem

Note: GitHub is now an official plugin (self-contained, auto-start) rather than an MCP server. It is enabled by default via enabledPlugins in settings.json and no longer requires manual token configuration in .mcp.json.

Agent Docs

Framework-specific reference in agent_docs/. Real content, not placeholder text:

• architecture.md - e.g., Next.js App Router with RSC/Client component boundaries
• conventions.md - naming, file structure, import ordering and error handling
• testing.md - test strategy, fixtures, mocking and coverage expectations
• deployment.md - deployment workflow and infrastructure patterns for your stack
• cache-friendly-workflow.md - practices for maximizing prompt cache hit rates

CLAUDE.md references these via @import syntax, so Claude Code auto-loads them without Read tool calls. They're still outside the cached prefix to keep token costs low.

---

Define Your Stack and Process

cc-rig separates what you're building (template) from how you like to work (workflow). Pick each independently. Any combination works.

Full showcase: 16 templates × 5 workflows × harness levels

What you're building: Templates

Templates set your language, framework, tool commands, linting and framework-specific rules:

Template	Stack	Highlights
generic	Language-agnostic	DevOps, monorepos, docs, infra — no framework assumptions
fastapi	Python + FastAPI	Async patterns, Pydantic, pytest, ruff
django	Python + Django	Fat models, ORM patterns, manage.py test
flask	Python + Flask	Blueprints, extensions, pytest, ruff
nextjs	TypeScript + Next.js	App Router, RSC patterns, Tailwind
gin	Go + Gin	Handler→Service→Repository, golangci-lint
echo	Go + Echo	Echo conventions, go test
go-std	Go (stdlib)	Idiomatic Go, no framework, go test, golangci-lint
rust-cli	Rust + Clap	CLI patterns, cargo test, clippy
rust-web	Rust + Axum	Async extractors, tower middleware, cargo test
rails	Ruby + Rails	MVC, ActiveRecord, minitest, rubocop
laravel	PHP + Laravel	MVC, Eloquent, Artisan, PHPUnit, PHP-CS-Fixer
express	Node.js + Express	Middleware patterns, Router, Jest, ESLint
phoenix	Elixir + Phoenix	LiveView, Ecto, ExUnit, Credo
spring	Java + Spring Boot	DI, JPA, JUnit 5, Checkstyle/Spotless
dotnet	C# + ASP.NET Core	DI, EF Core, xUnit, dotnet format

How you like to work: Workflows

Workflow	Best for
speedrun	Side projects, prototypes, hacking. 3 agents, 6 commands, 6 hooks. No memory. Just code fast.
standard	Most projects. 5 agents, 9 commands, 10 hooks. Memory, safety hooks, code review, architecture.
spec-driven	Teams that plan first. 9 agents, 13 commands, 11 hooks. Spec create/execute, PM and implementer agents.
gtd-lite	Task-oriented developers. 8 agents, 13 commands, 11 hooks. GTD capture/process, daily planning.
verify-heavy	Production systems. 13 agents, 19 commands, 14 hooks. Security auditor, PR reviewer, every quality gate.

Add-ons

Some workflows include compound features that span multiple Claude Code primitives:

Spec Workflow (spec-driven, verify-heavy). Plan-first development: /spec-create and /spec-execute commands, pm-spec and implementer agents, specs/TEMPLATE.md starter file. Based on Pimzino's spec workflow.

GTD System (gtd-lite). Getting Things Done for Claude Code: /gtd-capture, /gtd-process, /daily-plan commands, pre-created tasks/inbox.md, tasks/todo.md and tasks/someday.md. Based on adagradschool's cc-gtd.

Worktrees (spec-driven, gtd-lite, verify-heavy). Parallel development using Claude Code's native git worktree support: parallel-worker agent and /worktree command. For batch orchestration, cc-rig worktree spawn launches multiple Claude sessions in isolated worktrees simultaneously — each gets its own branch, runs independently, and can be merged via PR when done.

Mix and match

# Next.js prototype - move fast
cc-rig init --template nextjs --workflow speedrun

# FastAPI production API - maximum rigor
cc-rig init --template fastapi --workflow verify-heavy

# Go microservice - task-driven workflow
cc-rig init --template gin --workflow gtd-lite

---

For Teams

Save and share your config

Every cc-rig init saves a config file to your project. Commit it and teammates get the same setup:

# Teammate clones the repo, then:
cc-rig init --config .cc-rig.json

Same agents, same hooks, same permissions. Only project name and output directory change.

Export a portable config

Strip machine-specific paths for clean sharing:

cc-rig config save --export team.json --portable

Lock a config

Prevent modifications via expert mode. Teammates can still add custom CLAUDE.md rules (always additive), but can't change agents, hooks or commands:

cc-rig config lock my-app

Browse and compare

cc-rig config list              # See all saved configs
cc-rig config inspect my-setup  # View config details
cc-rig config diff my-setup     # Diff against current project

---

Going Deeper

Expert mode

Full control over everything: agents, commands, hooks, skills, MCP servers, permissions, features and custom CLAUDE.md rules. Starts from your workflow's defaults:

cc-rig init --expert

Autonomous mode

Claude works through a task list while you're away. Add a harness to any cc-rig project:

Harness options: from scaffold to autonomous loops

cc-rig harness init --lite        # Task tracking + session-start summary
cc-rig harness init               # + enforcement gates (lint blocks commits) + init-sh.sh
cc-rig harness init --autonomy    # + loop script, 5-step PROMPT.md, progress ledger

Each level builds on the previous. Or pick individual features à la carte — the wizard's "Custom" option lets you enable any combination of task tracking, budget awareness, verification gates and autonomy loop independently. A 6th option enables the ralph-loop plugin — Anthropic's official autonomous iteration loop plugin — as an alternative to cc-rig's generated loop.sh.

The standard level generates init-sh.sh (wraps your test/lint/format commands) and a commit-gate hook that structurally blocks commits when lint fails. The autonomy level generates loop.sh and PROMPT.md (5-step workflow: assess, advance, tidy, verify, record), an external bash loop that feeds tasks to Claude one at a time, each with fresh context. Based on the Ralph Wiggum technique by Geoffrey Huntley.

./loop.sh           # Run the autonomy loop (default: 20 iterations max)
./loop.sh 50        # Override max iterations

Warning: loop.sh uses --dangerously-skip-permissions. Run inside a Docker container or sandboxed environment. See Claude Code security docs.

Safety rails included: iteration limits, budget enforcement (stops the loop when token budget exceeded, warns at configurable threshold), checkpoint auto-commits (when Claude doesn't commit, the loop does), stuck detection, entropy management (tidy between iterations), per-iteration cost tracking in the progress ledger and a cost summary on exit (Ctrl+C, budget exceeded or completion).

The budget-reminder Stop hook shows actual session token usage and estimated cost (parsed from Claude Code's JSONL session logs) every time a session ends. Works at all harness levels (B1+), degrades gracefully when python3 is unavailable.

Health check and cleanup

Validate your configuration after use. Checks file integrity, hook permissions, memory files and manifest consistency:

cc-rig doctor                 # Check project health
cc-rig doctor --fix           # Auto-fix safe issues (permissions, missing files)

Remove everything cc-rig generated, using the manifest. Only touches what cc-rig created:

cc-rig clean

---

CLI Reference

cc-rig init [options]              Set up a new project
  --template <name>                Template preset (fastapi, nextjs, etc.)
  --workflow <name>                Workflow preset (standard, speedrun, etc.)
  --name <name>                    Project name
  -o, --output <dir>               Output directory
  --in-place                       Write to current directory
  --quick                          Quick picker (numbered lists)
  --expert                         Expert mode (full control)
  --migrate                        Detect and set up existing project
  --config <path>                  Load a saved config

cc-rig preset list [--templates|--workflows]
cc-rig preset inspect <name>       View preset details
cc-rig preset create <name>        Create preset from project config
cc-rig preset install <path>       Install a local preset file

cc-rig config save|load|list|inspect|diff|lock|unlock
cc-rig config update [-d DIR] [--quick|--expert]  Re-run wizard on existing config

cc-rig harness init [--lite|--standard|--autonomy] [-d DIR]
cc-rig harness status [--dir DIR]  Show current harness level and progress

cc-rig skills list [-d DIR]        Show installed skills
cc-rig skills catalog              Browse all available skills
cc-rig skills add <name> [-d DIR]  Install a skill from the catalog
cc-rig skills remove <name> [-d DIR] Remove an installed skill
cc-rig skills install [-d DIR]     Retry failed downloads from init

cc-rig worktree spawn "task1" "task2" ...  Launch Claude in parallel worktrees
cc-rig worktree list [-d DIR]      Show all worktrees and status
cc-rig worktree status <name>      Detailed status of one worktree
cc-rig worktree pr <name>          Push branch and create PR from worktree
cc-rig worktree cleanup [name] [--all] [--merged] [--force]

cc-rig doctor [--fix] [--check-compat] [-d DIR]  Validate project health
cc-rig clean [--force] [-d DIR]    Remove generated files

---

Workflow Philosophy

cc-rig's defaults encode seven workflow principles distilled from how the Claude Code team builds software, inspired by Boris Cherny's workflow (creator of Claude Code):

Principle	cc-rig Implementation
Plan before coding	/plan and /assumptions commands, /research for codebase exploration, CLAUDE.md workflow guidance
Use subagents for research	/research command, explorer agent (Haiku), parallel-worker for worktree isolation
Self-improvement loop	Auto-memory (personal), team memory (/remember, memory-precompact hook), persistent memory/ files
Verify before done	Hooks (format, lint, typecheck), B2+ enforcement gates (lint blocks commits), guardrails in CLAUDE.md
Demand elegance	/refactor command, refactorer agent, workflow principles in CLAUDE.md
Fix failures immediately	B1+ session-start task summary, B2+ commit-gate hook, B3 autonomy loop (stuck detection)
Track work with tasks	tasks/todo.md (B1+), GTD system (inbox/todo/someday), /daily-plan

Each workflow preset dials these principles up or down:

• speedrun — Minimal: just code fast. Verification hooks present but no process enforcement.
• standard — Core principles active. Plan, verify, remember, refactor.
• spec-driven — Plan-first emphasis. Specs before implementation.
• gtd-lite — Task management emphasis. Capture, process, plan.
• verify-heavy — All seven principles at maximum. Every quality gate active.

---

FAQ

How is this different from writing CLAUDE.md by hand?

You could write CLAUDE.md yourself. But a fully configured project also needs settings.json with hooks and permissions, agent markdown files with YAML frontmatter and tool restrictions, slash command files, skills, MCP config, memory files and agent docs, all with correct cross-references. cc-rig generates everything in seconds with content specific to your framework.

Does this work with existing projects?

Yes. cc-rig init --migrate scans your repo, detects your stack and proposes what to add. It only writes new files and won't touch anything that already exists.

Can I edit the generated files?

Yes. Everything is plain text. Edit whatever you want. cc-rig won't overwrite your changes. There's no "update" command. Generate once, own forever. For personal preferences, use CLAUDE.local.md (not git-tracked) to avoid conflicts with team config.

What about Claude Code plugins and skills?

cc-rig handles both community skills and official Anthropic plugins. For skills, cc-rig installs a starter set of community skills matched to your framework from 10 repos, with optional packs for deeper coverage. For plugins, cc-rig curates a 24-plugin official marketplace catalog across 5 categories (LSP, integration, workflow, autonomy, utility) and writes enabledPlugins into settings.json with smart defaults resolved by language, template and workflow. You can install any additional skill from skills.sh (73K+), awesome-claude-skills or any GitHub repo.

What's the autonomous mode?

A harness that lets Claude work through a task list unattended. It uses structural enforcement: hooks that block commits when lint fails, a utility script (init-sh.sh) wrapping your test/lint/format commands, iteration limits, budget enforcement with cost tracking, checkpoint auto-commits, stuck detection and an emergency stop. On exit you get a cost summary showing total tokens and estimated spend. Set it up, walk away, review the results.

Does this cost anything?

cc-rig is free and open source. Claude Code itself requires an Anthropic plan. cc-rig keeps CLAUDE.md lean and prompt cache hit rates high to minimize your token costs.

---

Community & Inspiration

Skill Ecosystem

cc-rig's starter set and optional packs draw from these repos. Skills are downloaded at init time from the original repos — cc-rig does not bundle or redistribute them.

• obra/superpowers - 14 SDLC workflow skills (MIT)
• trailofbits/skills - 30+ security + modern dev skills (CC-BY-SA-4.0)
• anthropics/skills - 16 official Anthropic skills (Apache 2.0)
• addyosmani/web-quality-skills - Core Web Vitals, accessibility, SEO (MIT)
• hashicorp/agent-skills - Official Terraform and Packer
• vercel-labs - React, Next.js and design guideline skills (MIT)
• supabase/agent-skills - PostgreSQL best practices (MIT)
• planetscale/database-skills - MySQL, PostgreSQL, Vitess (MIT)
• akin-ozer/cc-devops-skills - 31 CI/CD, IaC and monitoring skills (Apache 2.0)
• ahmedasmar/devops-claude-skills - Kubernetes, Terraform, monitoring, GitOps (MIT)
• agamm/claude-code-owasp - OWASP Top 10:2025 security (MIT)
• wshobson/agents - Tailwind CSS design system (MIT)
• peteromallet/desloppify - Code quality scoring and remediation (MIT)

The broader ecosystem has thousands more. Discover skills at skills.sh (73K+), awesome-claude-skills, skillsmp.com or install from any GitHub repo with npx skills add.

Research & Inspiration

cc-rig's defaults draw from community research on what makes Claude Code work well:

• Thariq on prompt caching - how Claude Code's team optimizes caching (informed cc-rig's cache-aware architecture)
• Boris Cherny's workflow principles - seven principles for effective Claude Code usage (creator of Claude Code)
• What great CLAUDE.md files have in common - content patterns
• HumanLayer's CLAUDE.md guide - structure principles
• Spec workflow by Pimzino - plan-first development
• cc-gtd by adagradschool - GTD for Claude Code
• SuperClaude - runtime framework, different approach
• QMD by Tobi Lütke - local doc search
• awesome-claude-code - community directory
• Ralph Wiggum technique by Geoffrey Huntley - autonomous loop pattern

---

Contributing

PRs welcome, especially new templates, new workflow presets, community skill integrations and bug fixes. Please open an issue first for large changes.

git clone https://github.com/runtimenoteslabs/cc-rig.git
cd cc-rig
python3 -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pytest tests/
ruff check cc_rig/

Zero runtime dependencies. Stdlib only. pytest and ruff are dev-only.

---

Ready to try it?
pip install cc-rig && cc-rig init

Star the repo if cc-rig saves you setup time.

---

License

MIT