Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
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
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Docs

ASecurity

Governs published docs (README.md, docs/, SETUP/TROUBLESHOOTING/CONTRIBUTING/SECURITY), including the reference-doc format for docs/commands.md, docs/configuration.md, docs/daemon-api.md. Load on any published-doc edit — create, change, move, trim, or link fix. Enforces open-source standards: Diataxis mode, granular single-topic files, caveman prose, one source per topic, link don't restate. Which facts land in the reference docs (config-doc rule) stays with AGENTS.md/CLAUDE.md and the spur s...

3 stars
0 votes
0 copies
1 views
Added 10/2/2026
documentationgogitapisecurity

Works with

cliapi

Security Analysis

A100/100

Scanned 10/2/2026

$npx -y skills add ashugaev/spur --skill docs --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Docs?

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

Security grade badge for Docs
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/ashugaev-docs/badge)](https://www.skillsdirectory.com/skills/ashugaev-docs)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: docs
description: "Governs published docs (README.md, docs/, SETUP/TROUBLESHOOTING/CONTRIBUTING/SECURITY), including the reference-doc format for docs/commands.md, docs/configuration.md, docs/daemon-api.md. Load on any published-doc edit — create, change, move, trim, or link fix. Enforces open-source standards: Diataxis mode, granular single-topic files, caveman prose, one source per topic, link don't restate. Which facts land in the reference docs (config-doc rule) stays with AGENTS.md/CLAUDE.md and the spur skill; SKILL.md/agent prose belongs to skill-writer."
---

DOCS MANAGEMENT: root README is a minimal hub, Docs section lists every
published doc plus the AI-agent install prompt; reference lives in docs/.

WRITE CAVEMAN

  Cut what a mid dev or agent already knows (npm, systemd, git, nginx
  basics); state the gotcha, skip the step. No filler, self-narration
  ("as noted", "simply", "note that"), marketing, future promises, time words
  ("now", "currently"). Fragments OK, imperative, one term per concept.

COMPACT, HARD

  One fact per line. Word over clause, clause over sentence, sentence
  over paragraph.
  Second pass on every block you touch: cut 30% of the words, lose zero
  facts. Merge two sentences sharing a subject.
  Paragraph over 4 lines: split into a `-` list or drop the narration.
  Word bans are the CAVEMAN, HARD list in `skill-writer` — hedge, degree,
  meta, justifying, transition, courtesy. One hit is a defect.
  Compression never drops a default, a limit, a field name, or an event
  name. Numbers and identifiers stay byte-exact.

SCOPE: document the interface, never the change

  Document user-facing surface: command, flag, config field, source type,
  event, provider, API route, install/deploy/CLI behavior.
  No doc edit for a refactor, an internal helper, a perf fix, or a bug fix
  that leaves the interface unchanged. Not every commit earns a doc line.
  State what a field does, its default, its constraint. Never why it was
  built, which design lost, or what the code does internally.
  Internal mechanism — cache keys, tick order, call chains, data shapes —
  stays in code, unless an operator acts on it.

GRANULAR: one doc, one topic, one Diataxis mode; split on drift.

  Tutorial (first run)          docs/install-from-npm.md
  How-to (one task)             docs/install-from-source.md, docs/voice.md,
                                 TROUBLESHOOTING.md, SETUP.md
  Reference (commands, fields)  docs/commands.md, docs/configuration.md,
                                 docs/daemon-api.md
  Explanation (why)             gitignored planning notes, not shipped

REFERENCE DOCS: docs/commands.md, docs/configuration.md, docs/daemon-api.md are
navigation for an agent that reads source freely, never a copy of the source.

  One line per surface item: name, default only if verified against current
  source, pointer to the owning file for behavior (`v2/src/config.ts`,
  `v2/src/cli.ts`, a route handler, a source file under `v2/src/event-sources/`).
  No sub-bullets, no multi-sentence entries. Sibling keys sharing an owner
  collapse onto one line (`ports.<id>.{env,start,end,url}: ... file.ts`).
  Banned: mechanism (tick order, call chains, race handling, exact algorithm
  steps), rationale, history, an example past one minimal skeleton, restating
  `--help` output or a code constant a source file defines, and any
  multi-clause sentence describing behavior — that belongs to code and its
  own tests, never to this file.
  New surface ships with exactly one line added. Nothing else grows.
  Line budget per file, hard cap: commands.md 132, configuration.md 250,
  daemon-api.md 94. daemon-api.md only: one method + one full path per
  line, always — never `GET|POST /x` or `/x/a|b`, a route grep must find
  its exact `METHOD /path` string verbatim on one line. Over budget on your
  own edit: cut elsewhere in the same file before adding.
  A section that is a guide or how-to, not a lookup table: drop it. Check
  inbound anchors first (`git grep -n '\.md#<anchor>'`) — a live inbound
  anchor keeps a one-line heading, everything else in the section goes.
  Never pin a test's assertion to a doc's exact sentence. A test verifies a
  surface exists and is findable (name, flag, route, event, status code);
  it never re-derives behavior from doc prose.

ONE SOURCE: each topic has one owning doc, everything else links.

  - Commands: docs/commands.md only. Config fields: docs/configuration.md only.
    Route names: docs/daemon-api.md only — full payload/status-code contracts
    live in the handler (`v2/src/server.ts`), per daemon-api.md's own scope line.
    A route documented as part of a config field's behavior stays with that field.
  - Contradiction: pick the owner, delete the copy, link.

LINK GRAPH: every published doc reachable from the README Docs section, relative links only.

  Rename or move updates every inbound reference: markdown links, the .claude/.agents
  skill trees, CI path filters in .github/workflows/*.yml — link check
  misses the last two. Planning notes stay gitignored.

CHECKLIST

  - Nothing a mid dev or agent already knows
  - No interface change, no doc change
  - Second compression pass done
  - One mode, one topic per doc
  - One owner per topic; duplicates are links
  - Reachable from the README.md Docs section, or intentionally gitignored
  - No dead relative links
  - Numbers and commands match the canonical source
  - Reviewer treats undocumented new surface as a defect

BOUNDARIES: docs/commands.md / docs/configuration.md / docs/daemon-api.md
plus the spur skill own command, config field, and daemon route semantics;
SKILL.md, agent definitions, AGENTS.md/CLAUDE.md prose belong to skill-writer.

Attribution

ashugaevashugaev
View sourceSee grades on GitHubMore from ashugaev →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

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

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Context Fundamentals

Understand the components, mechanics, and constraints of context in agent systems. Use when designing agent architectures, debugging context-related failures, or optimizing context usage.

179001 votes

Architecture Diagram Creator

Create comprehensive HTML architecture diagrams with data flows, business context, and system architecture.

6661 votes

release-notes

Draft release notes and changelog entries from git history or merged PRs between two refs (tags/SHAs/branches), including breaking changes, migrations, and upgrade steps. Use when the user asks for release notes, changelog updates, or a GitHub Release draft.

1301 votes

docs-style-guide

Documentation style guide enforcer by @planetabhi. Applies and reviews the writing style guide when authoring or editing product documentation and tutorials. Use to check prose for voice, tense, word choice, inclusive language, formatting, code block, UI, Markdown, and number/date conventions.

11 votes

Docx

Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', '.dotx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in W...

1798860 votes
View all in documentation →