CtrlK
BlogDocsLog inGet started
Tessl Logo

common-documentation

Write effective code comments, READMEs, and technical documentation following intent-first principles. Use when adding comments, writing docstrings, creating READMEs, or updating any documentation.

68

1.14x
Quality

75%

Does it follow best practices?

Impact

100%

1.14x

Average score across 1 eval scenario

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.github/skills/common/common-documentation/SKILL.md
SKILL.md
Quality
Evals
Security

Documentation Standards

Priority: P2 (MEDIUM)

1. Intent-First Comments

  • Explain "Why" logic exists. Avoid "What" mechanics.
  • Use triple-slash (Dart/Swift) or JSDoc (TS/JS) for public members.
  • Delete commented-out code. Use Git history.
  • Format: TODO(username): description. Link tickets.
  • For retry backoff, explain that jitter prevents the thundering herd; document intent rather than restating the delay calculation.

2. README Structure

  • Mission: Project purpose (one sentence).
  • Onboarding: Prerequisites, installation, usage (exact).
  • Maintenance: Document inputs/outputs, known quirks, fixes.
  • Sync: Documentation ships with feature.

3. ADRs & Architecture

  • ADRs: Document rationale for system changes in docs/adr/.
  • Docstrings: Include Args, Returns, and Usage examples (>>>).
  • Diagrams: Use Mermaid.js inside Markdown.

4. API Docs

  • Use Swagger/OpenAPI for REST.
  • Provide copy-pasteable examples for endpoints.
  • Define contract before implementation.

Anti-Patterns

  • No "what" comments: Explain intent. Refactor mechanics.
  • No orphan TODOs: Require owner and ticket.
  • No stale docs: Document during development.

Canonical response anchors

When this skill applies, preserve the following domain terminology or equivalent concrete examples in the answer when relevant:

  • JSDoc
  • Usage
  • why
Repository
HoangNguyen0403/agent-skills-standard
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.