planning-with-files
Optional persistent file-based planning for multi-phase or long-running work that benefits from recovery across sessions or…
Write or revise technical documentation, procedures, comments, commits, releases, and incident reports.
$ npx -y skills add khaneliman/khanelinix --skill technical-writing --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/technical-writingContext preview
The summary Claude sees to decide when to auto-load this skill.
Write or revise technical documentation, procedures, comments, commits, releases, and incident reports.
name: technical-writing description: Write or revise technical documentation, procedures, comments, commits, releases, and incident reports.
Make text easier to understand without deleting technical content. Treat source facts and structure as invariants, not optional detail. Claim full ASD-STE100 conformance only with the official current standard and controlled dictionary.
intended reader and text type.
protected content, rewrite, then score source against candidate.
[rules.md](references/rules.md) and obey repository contributor canon first.
uncertainty, impact, response, and follow-up ownership while applying the same sentence rules.
1. Identify audience, purpose, and procedural or descriptive mode. 2. Record facts, caveats, figures, identifiers, code, links, and tables that must remain. 3. Rewrite by clarifying sentences, naming actors, stabilizing terms, and removing filler. Never gain brevity by deleting protected content. 4. Run the bundled scorer:
python3 scripts/style_guard.py score SOURCE CANDIDATE --mode descriptive
5. Treat lexical retention as a measurable rewrite gate, not proof of fact retention. Resolve every missing number, code span, URL, table, or required fact. 6. Compare source and candidate directly. Confirm that meaning, scope, uncertainty, and causal claims did not change.
Use `--required-facts FILE` for one exact required item per line. Use `--minimum-retention` only to tune the lexical warning threshold.
`scripts/style_guard.py` provides three read-only checks:
The checks can reject measurable violations. Model judgment still owns factual equivalence, correct terminology, audience fit, and useful structure. Sentence length and ordinary word choice are contextual judgments, not rejection rules.
Run packages directly with: Or install from the packages output. For example:
Optional persistent file-based planning for multi-phase or long-running work that benefits from recovery across sessions or…
Audit coding-agent sessions and mine cross-harness history for recurring corrections, skill-routing gaps, and harness faults. Use…
Design and audit agent instructions, skills, plugins, subagents, prompts, hooks, and provider configuration. Use for AI-tool…
Sketch interfaces, state, and module structure before code. Use for explicit design-led implementation or as a caller-owned…
Spawn N parallel candidates at the same task, pick a base, graft the strongest parts of the losers into it. Use for /arena,…
Bevy workflows for ECS, migrations, first-party UI/input/rendering/assets/shaders, plugin selection, scenes, scheduling, hot…