project-local-skills
How to create, manage, and share project-local skills — .clarc/skills/ structure, scope hierarchy, team workflow, and promotion to global clarc
Best use case
project-local-skills is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
How to create, manage, and share project-local skills — .clarc/skills/ structure, scope hierarchy, team workflow, and promotion to global clarc
Teams using project-local-skills should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/project-local-skills/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How project-local-skills Compares
| Feature / Agent | project-local-skills | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
How to create, manage, and share project-local skills — .clarc/skills/ structure, scope hierarchy, team workflow, and promotion to global clarc
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Project-Local Skills
Project-local skills live in `.clarc/skills/` inside a project repository.
They are automatically loaded at session start and override global clarc skills of the same name.
## When to Activate
Use project-local skills when:
- A team has project-specific conventions (internal API patterns, company auth library)
- Conventions differ from clarc's global rules for this codebase
- A migration is underway and the pattern is temporary
- The skill should be versioned alongside the code it describes
## Directory Structure
```
<project-root>/
.clarc/
skills/
our-auth-pattern/
SKILL.md ← project-local skill
internal-api/
SKILL.md
instincts.md ← personal learned patterns (gitignore this)
context.md ← last session context
progress.md ← current progress
```
## Scope Hierarchy
Skills are resolved in this order (highest precedence first):
```
1. .clarc/skills/<name>/ → project-local (this project only)
2. ~/.clarc/skills/<name>/ → global clarc (all projects)
3. Claude Code native skills → built-in system-reminder skills
```
When a local skill overrides a global one, session-start logs:
`Local skill 'typescript-patterns' overrides global version`
## Creating a Project-Local Skill
```bash
/skill-create --local <name>
```
Scaffolds `.clarc/skills/<name>/SKILL.md`:
```markdown
---
title: <Name>
scope: project-local
tags: []
created: <date>
team:
---
## When to Use
<situation that calls for this skill>
## Pattern
<the approved approach, with code examples>
## Anti-patterns
<what NOT to do>
## References
<internal docs, ADRs, tickets>
```
Fill in the sections, then commit `.clarc/skills/` to the repo.
## Team Workflow
```
1. Developer creates skill: /skill-create --local our-auth-pattern
2. Fills in pattern content
3. Commits .clarc/skills/ to the project repo
4. Team pulls → all members get the skill at next session start
5. When ready for global use: /promote-skill our-auth-pattern
```
## Listing Local Skills
```bash
/skills-local
```
Output:
```
Project-local skills — .clarc/skills/ (2 skills)
our-auth-pattern How we use the internal auth library
payment-flow Approved payment integration pattern
Total: 2 local · 0 override global
```
## Promoting to Global
When a project-local skill is useful across multiple projects:
```bash
/promote-skill our-auth-pattern # Copy to ~/.clarc/skills/
/promote-skill our-auth-pattern --pr # Also open upstream PR
```
Safety checklist before promoting:
- [ ] Remove internal URLs and company-specific names
- [ ] Replace hardcoded paths with generic examples
- [ ] Ensure examples work without internal dependencies
## What to Gitignore
```gitignore
# Keep skills committed (shared with team)
# .clarc/skills/ ← do NOT ignore
# Keep instincts private (personal learning)
.clarc/instincts.md
.clarc/context.md
.clarc/progress.md
```
Or commit everything if the full memory bank should be shared.
## Anti-patterns
- Creating a local skill for something already covered by a global clarc skill (use `/find-skill` first)
- Promoting a skill that still contains internal URLs or company names
- Putting project-local skills in `~/.clarc/skills/` — those become global for all projects
- Forgetting to commit `.clarc/skills/` — other team members won't see the skill
## Related Commands
- `/skill-create --local <name>` — scaffold a new project-local skill
- `/skills-local` — list all local skills and override status
- `/promote-skill <name>` — promote a local skill to global scope
- `/find-skill <topic>` — search global clarc skills before creating a new oneRelated Skills
project-guidelines-example
Example project-specific skill template based on a real production application.
zero-trust-patterns
Zero-Trust security patterns — mTLS between microservices (Istio/SPIFFE), SPIRE workload identity, OPA/Envoy authorization, NetworkPolicy default-deny-all, short-lived credentials, service mesh security, and Kubernetes RBAC hardening.
wireframing
Wireframing and prototyping workflow: fidelity levels (lo-fi sketch → mid-fi wireframe → hi-fi prototype), tool selection (Figma, Excalidraw, Balsamiq), user flow diagrams, wireframe annotation standards, information architecture (IA) mapping, and the handoff from wireframe to visual design. For developers who need to communicate UI structure before writing code.
webrtc-patterns
WebRTC patterns — peer connection setup, ICE/STUN/TURN configuration, signaling server design, SFU vs mesh topology, screen sharing, media track management, and reconnect/ICE restart handling.
webhook-patterns
Webhook patterns for receiving, verifying (HMAC), and idempotently processing third-party events. Covers Stripe, GitHub, and generic webhook patterns, delivery guarantees, retry handling, and testing.
web-performance
Web performance optimization: Core Web Vitals (LCP, CLS, INP), Lighthouse CI with budget configuration, bundle analysis (webpack-bundle-analyzer, vite-bundle-visualizer), hydration performance, network waterfall reading, image optimization (WebP/AVIF, srcset), and font performance.
wasm-performance
WebAssembly performance: wasm-opt binary optimization, size reduction (panic=abort, LTO, strip), profiling WASM in Chrome DevTools, memory management (linear memory, avoiding GC pressure), SIMD, and multi-threading with SharedArrayBuffer.
wasm-patterns
WebAssembly patterns: wasm-pack, wasm-bindgen (JS↔Wasm interop), WASI, Component Model, wasm-opt, Rust-to-WASM compilation, JS integration (web workers, streaming instantiation), and production deployment (CDN, Content-Type headers).
visual-testing
Visual Regression Testing: tool comparison (Chromatic/Percy/Playwright screenshots/BackstopJS), pixel-diff vs AI-based comparison, baseline management, flakiness strategies (masks, tolerances, waitForLoadState), CI integration with GitHub Actions, and Storybook integration.
visual-identity
Brand identity development: color palette construction (primary/secondary/semantic/neutral), logo concept brief writing, typeface pairings, brand voice definition, mood board direction, and Brand Guidelines document structure. Use when establishing or evolving a visual brand — not for implementing existing tokens.
ux-micro-patterns
UX micro-patterns for every product state: Empty States, Loading States (skeleton screens, spinners, optimistic UI), Error States, Success States, Confirmation Dialogs, Onboarding Flows, and Progressive Disclosure. These patterns apply to every feature — done wrong, they're the biggest source of user confusion.
typography-design
Typography as a creative discipline: typeface selection criteria, type pairing (serif + sans, display + body), modular scale systems, line-height and tracking ratios, hierarchy construction, and web/mobile rendering considerations. The decisions behind design tokens, not the tokens themselves.