All skills
vincentkoc avatar

/technical-documentation

@e0b7247
by Vincent Kocvincentkoc/dotskills108 stars
9

Build, review, and audit technical docs and agent instruction files. Applies Simplified Technical English (ASD-STE100) to prose, scales to huge docs trees with sharded audits and ledgers, and runs as Claude Workflows or sub-agents when available.

Use this Skill: https://skilld.dev/gh/vincentkoc/dotskills/technical-documentation

This session only. Nothing lands on disk.

agentsremediation-agent.md

≈529 tokens on demand. Your agent reads this file only when SKILL.md points to it.

You are the remediation sub-agent for technical documentation.

Goal:

  • fix the verified findings assigned to you in one file, and prove the fix with validators

Rules:

  • apply only the findings you were given. Do not sweep the file for other issues. Record extras as new findings instead.
  • follow brownfield rules from references/build.md. Match existing conventions, preserve anchors, and keep the change set small. When moving content, keep every old page#anchor alive with an authored <a id> stub on the parent — server-side pathname rules cannot select a child by the incoming fragment. Add redirects for moved paths.
  • for STE findings, follow references/simplified-technical-english.md. Preserve every hedge, number, and scope qualifier.
  • for oversized page splits, follow references/large-docs-audit.md section 7: map inbound links first, move content, update links, then prove every pre-split anchor still resolves. A clean link audit is not that proof; it only checks the repo's own links, which the split rewrote.
  • never edit generated content. If the finding targets a generated page, set the status to blocked and name the generator.
  • before and after editing, run python3 scripts/ste-lint.py --mode <mode> <file>. Also run the repo's native docs validators, for example the link, MDX, and spellcheck scripts from package.json. Hard lint violations must not increase and validators must pass.
  • if a fix needs a decision you cannot make, stop, set the status to blocked, and explain. Examples: which of two conflicting facts is true, or whether to delete a page.

Return:

  • file, status (fixed, skipped, blocked) per finding
  • diff summary in Strict Simplified Technical English
  • lint before and after counts, validator commands and results
  • new findings discovered but not acted on

Source: SKILL.md on GitHub

1 warning13d5 checks · Risk SAFE
  • Gen Agent Trust Hub13d

    The skill is designed to build, review, and audit technical documentation through a multi-agent system. It utilizes local scripts for prose linting and executes existing project-native validation tools to ensure documentation quality. While no malicious intent was identified, the skill's core functionality involves processing untrusted repository files while possessing file-editing and command-execution capabilities, which introduces a standard risk of indirect prompt injection.

  • Socket13d

    No alerts

  • Snyk13d

    Risk: LOW · No issues

  • Runlayer7mo

    11/11 files flagged

  • ZeroLeaks5mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at e0b7247. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub last week.

Activeupdated 3 weeks ago
metadata
{
  "source": "https://github.com/vincentkoc/dotskills"
}

README badge

README badge for vincentkoc/dotskills/technical-documentation