cli-code-skills

cli-code-skills
Drop-in Claude Code skills that audit your code and generate your docs, designs, and infra.

cli- = Command Line Interface + Clement Liard Initials

Quickstart

git clone https://github.com/Destynova2/cli-code-skills.git
cli-code-skills/scripts/install_skills.sh

Type /cli- in Claude Code or Codex, hit tab, pick a skill:

/cli-audit-code src/          → Clean Code score with per-category breakdown
/cli-audit-xray src/core/     → Semantic optimization cards with invariants and validation
/cli-audit-test tests/        → Test plan quality score (ISTQB/TMMi, 12 dimensions)
/cli-forge-hld "payment API"  → HLD with C4 L1-L2, capacity estimation, ADRs
/cli-forge-lld "order service" → LLD with class/sequence diagrams, API specs, DB schemas
/cli-forge-schema "auth flow" → GitHub-compatible Mermaid diagram, auto-split if complex
/cli-cycle .                  → Full project review with prioritized action plan

Skills

cli-cycle — Orchestrator

cli-audit-* — Analyze & Score

cli-forge-* — Generate

Other Skills

Naming Convention

cli-<action>-<target>
 │    │        │
 │    │        └─ what it operates on (code, doc, hld, lld, infra, readme, schema, tree, test)
 │    └────────── what it does (audit = analyze, forge = generate, cycle = orchestrate)
 └─────────────── namespace

Installation

Claude Code

git clone https://github.com/Destynova2/cli-code-skills.git
cli-code-skills/scripts/install_skills.sh --claude

Codex

git clone https://github.com/Destynova2/cli-code-skills.git
cli-code-skills/scripts/install_skills.sh --codex

Cursor IDE

git clone https://github.com/Destynova2/cli-code-skills.git
cp -r cli-code-skills/.cursor/rules/* your-project/.cursor/rules/

Update

cd cli-code-skills && git pull
scripts/install_skills.sh

Validation

Run the repository checks before publishing skill changes:

python3 scripts/validate_skills.py

The validator checks the skill count, frontmatter, README/CLAUDE inventories,
shared-file paths, and Claude/Codex-neutral cli-cycle orchestration. The same
check runs in GitHub Actions on pushes and pull requests.

Project Structure

cli-code-skills/
├── .github/workflows/     # Repository validation workflow
├── .claude-plugin/        # Plugin manifest for marketplace
│   └── plugin.json
├── .cursor/rules/         # Cursor IDE rules (4 core skills)
├── docs/reviews/          # cli-cycle audit reports
├── cli-cycle/             # /cli-cycle — orchestrator
├── cli-audit-code/        # /cli-audit-code — Clean Code scoring (CQI)
│   └── references/        #   12 categories, scoring framework, anti-patterns
├── cli-audit-xray/        # /cli-audit-xray — semantic optimization scanner
│   └── references/        #   SOG, optimization families, examples, risk rules
├── cli-audit-doc/         # /cli-audit-doc — documentation quality (DQI)
│   └── references/        #   12 categories, scoring framework, anti-patterns
├── cli-audit-drift/       # /cli-audit-drift — semantic drift against contracts
│   └── references/        #   contract templates and drift detection rules
├── cli-audit-review/      # /cli-audit-review — strict MR/PR review scoring
│   └── references/        #   RMI scoring, conventions, comments, and methodology
├── cli-audit-shell/       # /cli-audit-shell — shell quality and safety
│   └── references/        #   shell categories, scoring, anti-patterns
├── cli-audit-sync/        # /cli-audit-sync — doc-code coherence
│   └── references/        #   coherence layers and duplication rules
├── cli-audit-tangle/      # /cli-audit-tangle — call graph topology
│   └── references/        #   graph methods, analogies, fix patterns
├── cli-audit-test/        # /cli-audit-test — test plan quality & maturity
│   └── references/        #   dimensions, techniques, anti-patterns, scoring
├── cli-audit-wizard/      # /cli-audit-wizard — setup/config UX lifecycle
│   └── references/        #   flows, doctor checks, tooling ladder
├── cli-forge-arch/        # /cli-forge-arch — HLD/LLD router
│   └── references/        #   templates (backward compat)
├── cli-forge-chef/        # /cli-forge-chef — multi-agent brigade orchestrator
│   └── references/        #   prompts, tmuxinator, quality gates
├── cli-forge-choice-ux/   # /cli-forge-choice-ux — choice, checkout, signup, settings UX
│   └── references/        #   principles, scorecard, flow patterns
├── cli-forge-demo/        # /cli-forge-demo — reproducible live-demo kit
│   └── references/        #   staging, dramaturgy, tooling, templates
├── cli-forge-github/      # /cli-forge-github — GitHub repository health
│   └── references/        #   repo health patterns and checks
├── cli-forge-hld/         # /cli-forge-hld — High-Level Design
│   └── references/        #   sections, scoring, anti-patterns, estimation
├── cli-forge-doc/         # /cli-forge-doc — full project documentation
│   └── references/        #   Diataxis framework templates
├── cli-forge-infra/       # /cli-forge-infra — ops integration
│   └── references/        #   scoring, patterns, checklists
├── cli-forge-lld/         # /cli-forge-lld — Low-Level Design
│   └── references/        #   sections, scoring, anti-patterns, LLD template
├── cli-forge-oci-rootless/ # /cli-forge-oci-rootless — rootless OCI migration
│   └── references/        #   scenarios, proof gates, handoffs
├── cli-forge-perf/        # /cli-forge-perf — performance optimization method
│   └── references/        #   profiling, benchmarks, domains, examples
├── cli-forge-pipeline/    # /cli-forge-pipeline — CI/CD pipeline optimizer (biomimetic)
│   └── references/        #   pipeline templates, scoring, cache strategy
├── cli-forge-prez/        # /cli-forge-prez — technical slide decks
│   └── references/        #   slide structure and presentation patterns
├── cli-forge-quorum/      # /cli-forge-quorum — REC-Quorum orchestration
│   └── references/        #   quorum phases, templates, gotchas
├── cli-forge-readme/      # /cli-forge-readme — README generation (RCI)
│   └── references/        #   README pyramid and project types
├── cli-forge-resilience/  # /cli-forge-resilience — resilience, runbooks, agent-ready ops
│   └── references/        #   models, tests, runbook templates, agent ops
├── cli-forge-schema/      # /cli-forge-schema — Mermaid diagrams
│   └── references/        #   diagram types, conversions, modes
├── cli-forge-tree/        # /cli-forge-tree — directory structure (SHS)
│   └── references/        #   archetypes, conventions
├── cli-git-conventional/  # /cli-git-conventional — commits, SemVer, branches
│   └── references/        #   branch, command, and SemVer rules
├── cli-watermark/         # /cli-watermark — code provenance watermarking
│   └── references/        #   watermark layers and proof bundle
├── shared/                # Shared cross-skill contracts
├── scripts/               # Installation and repository validation utilities
├── gotchas.md             # Persistent lessons-learned, read by all skills
└── README.md

Each skill is a self-contained directory:

cli-<skill>/
├── SKILL.md          # Prompt: workflow, checklist, principles
└── references/       # Domain knowledge: templates, standards, cheatsheets

Skills run as forked agents with scoped tool access. They read your code, produce structured markdown with scores, diagrams, and actionable next steps.

Contributing

PRs welcome. To add a skill:

1. Create a directory following cli-<action>-<target>
2. Add a SKILL.md with frontmatter (name, description, context: fork)
3. Optionally add references/ for domain knowledge
4. Submit a PR

License

MIT