Skip to content
Back to skills

Technical Writer

ASecurity

<!-- SKILL.md — Technical Writer operating manual. Runtime: Claude Code + ACP (swappable coding CLI). Content stays LCD — no assumptions about a specific stack or project. -->

  • 3 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 20, 2026
ai-agentsgitapi

Works with

  • claude code
  • cli
  • api

Security analysis

A100/100

Pro scans all 7 files and shows the line behind each finding

Scanned October 20, 2026

npx -y skills add oneafrikan/the-grid --skill technical-writer --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Technical Writer?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Technical Writer
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/oneafrikan-technical-writer/badge)](https://www.skillsdirectory.com/skills/oneafrikan-technical-writer)

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

Download with Pro
SKILL.md
<!--
  SKILL.md — Technical Writer operating manual.
  Runtime: Claude Code + ACP (swappable coding CLI). Content stays LCD —
  no assumptions about a specific stack or project.
-->

# Skill: Technical Writer

## Invocation

```
/technical_writer <documents and/or the change to document>
```

Or picked up from a Signal Protocol entry / PR assigned to technical-writer. With
no scope given, audit the repo's documents against its code.

---

## Step 1 — List documents and their sources of truth

Pair each document with what it must be verified against:

| Document | Verify against |
|---|---|
| README (setup, usage) | entry points, package manifests, scripts, actually running the commands |
| CLAUDE.md / agent guidance | code layout, scripts, test and build config |
| API table / reference | route definitions, handlers, schemas |
| Env var reference | where the code or compose file reads each variable, `.env.example` |
| Command reference | scripts, Makefile, CLI definitions, `--help` output |
| Architecture / onboarding | actual modules, services, compose/deploy config |
| Changelog | git history, tags |
| Agent briefs (`.claude/agents`, skills) | the tools, paths and commands they cite |

Also list hand-synced mirrors (copies kept in sync by hand) by name.

## Step 2 — Read everything fully

- Read each document end to end, then the code it describes.
- Read the real code, compose, and config; do not take another document's word.
- Run commands where feasible.

## Step 3 — Build the discrepancy list

For each mismatch record: document, claim, what the code shows, fix.

- Check status words in both directions: "planned" that now exists, "exists" that does not.
- Check counts, names, paths, flags, defaults, versions.
- If the code looks wrong instead of the doc, mark it for the owning dev role; no doc edit.

## Step 4 — Fix with minimal edits

- Edit every affected document in the same pass, mirrors included.
- Change only the wrong sentence, row, or command; keep each document's voice and structure.
- Add no sections for things that do not exist.
- Re-run edited commands where feasible.

## Step 5 — Report

- One line per fix: file, what changed, what it was verified against.
- List code defects flagged, with the dev role they route to.
- If everything already matched, state that and make no edits.

---

## Checklist

- [ ] Every claim checked against code, not another doc
- [ ] "Planned" and "exists" correct in both directions
- [ ] All affected documents and mirrors updated together
- [ ] Commands copy-pasteable and run where feasible
- [ ] No cosmetic edits, no speculative sections

Files in this skill

  • AGENTS.md2.8 KB
  • IDENTITY.md581 B
  • MEMORY.md987 B
  • SKILL.md2.6 KB
  • SOUL.md3.1 KB
  • USER.md592 B
  • role.yaml1018 B

Attribution

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

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.