Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Documentation

ASecurity

Writes documentation for one reader doing one task: READMEs, quickstarts, how-to guides, API reference, runbooks, and changelogs. Use when docs are missing, stale, or unreadable, when a change made existing docs wrong, or when someone new cannot get the project running. Verifies every command and example before publishing.

9 stars
0 votes
0 copies
0 views
Added 9/27/2026
ai-agentsrustgoapidocumentation

Works with

api

Security Analysis

A100/100

Scanned 9/27/2026

Install to Claude Code

$npx -y skills add khuynh22/agent-dev-team --skill documentation --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Documentation?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Documentation
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/khuynh22-documentation/badge)](https://www.skillsdirectory.com/skills/khuynh22-documentation)

More formats (shields.io, HTML) on the badges page.

Download with Pro
Files
SKILL.md
---
name: documentation
description: Writes documentation for one reader doing one task: READMEs, quickstarts, how-to guides, API reference, runbooks, and changelogs. Use when docs are missing, stale, or unreadable, when a change made existing docs wrong, or when someone new cannot get the project running. Verifies every command and example before publishing.
license: MIT
metadata:
  phase: ship
  owners: [docs-engineer]
  version: "0.1.0"
---

# Documentation

Every document has one reader and one task. A document that serves everyone serves nobody,
and a feature tour is not documentation.

## Pick the type first

| Reader wants | Type | Shape |
|--------------|------|-------|
| To get it working now | Quickstart | Numbered steps, one path, copy-pasteable, ends in a verifiable result |
| To do a specific task | How-to | Goal, prerequisites, steps, verification |
| To look up an exact detail | Reference | Complete and uniform. Not a narrative |
| To understand why | Explanation / ADR | Prose, trade-offs, what was rejected |
| To fix production at 3am | Runbook | Symptom, first check, mitigation, escalation |

Mixing two types is the most common documentation failure. A quickstart with a paragraph
of rationale in the middle loses the reader who came to paste a command.

## Rules

1. **Name the reader and the task in the first two sentences.** If you cannot, the document
   has no scope yet.
2. **Every command is complete and copy-pasteable.** No placeholder the reader has to
   guess; if a value is required, say where to get it.
3. **Every procedure ends with how to know it worked**, with the expected output quoted.
4. **Show the common failure too.** The error someone will hit, and what it means, saves
   more time than the happy path.
5. **Link, do not duplicate.** Duplicated content goes stale in one place and stays correct
   in the other, which is worse than missing.
6. **Delete aggressively.** Wrong documentation is worse than none, because it is believed.
   If something is stale and you cannot verify it, remove it and say you did.
7. **Shortest version that works.** Length is a cost the reader pays.

## README order

Title, then one line saying what this is. Then: the problem it solves, install, a minimal
working example, links to deeper docs, how to contribute. Badges and a feature list do not
go above the example.

The example is the most-read part of any README. Make it real, small, and runnable as
written, including its imports.

## Runbook

```markdown
# Alert: <name>
**Means:** <the user-visible symptom>
**Check first:** <one dashboard or query>
**Common causes:** <ranked, each with its distinguishing signal>
**Mitigation:** <the action that stops the bleeding>
**Escalate to:** <who, and when>
```

Written for someone woken up with five minutes. No background, no history.

## Changelog

Grouped by version, newest first. Entries say what changed for a user, not what changed in
the diff. Breaking changes get their own section with one migration line each.

## Code comments and docstrings

Default to none. The bar is not "is this true" but "would a competent reader be stuck
without it".

Worth writing: a non-obvious *why* whose absence invites someone to "fix" the code and
break it; a cited constraint with its source; a deliberate deviation from convention.

Not worth writing: a restatement of the next line, a section banner, a docstring that
repeats the signature, or a note about the change you just made. Before reaching for a
comment, try a better name, a smaller function, or an earlier return.

## Verification

- Every command in the document was actually run, and the quoted output is real.
- Every code example runs as written, including imports.
- Every link resolves.
- Nothing contradicts the current code.
- The document is one type, not two.

## Red flags

| Thought | Reality |
|---------|---------|
| "I'll document what it should do" | Document what it does. Aspirational docs are wrong docs. |
| "The code is self-documenting" | Then the reference is cheap. The why is still missing. |
| "I'll leave the old section, it might still be right" | Might-be-right is believed. Verify or delete. |
| "More detail is more helpful" | Length is a cost. Cut what this reader does not need. |
| "The example is close enough" | Run it. A broken example destroys trust in the whole document. |
| "I'll add a comment explaining this function" | Try a better name first. |

Attribution

khuynh22khuynh22
View sourceMore from khuynh22 →
SSkills DirectorySkills Directory

Know which skills are safe — weekly.

Best new skills + every skill we flagged as malicious. From the team that scanned 103,619.

Join free

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Know which skills are safe — weekly.

Best new skills + every skill we flagged as malicious. From the team that scanned 103,619.

Join free

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

694821 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →