beat-writer — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited beat-writer (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Channel-agnostic, beats-first writer for uncommitted drafts. Owns extraction + shaping; delegates prose to /authors-voice (operator's default anchor — personal voice). Output is a plain OpenWriter doc that can stay generic OR get refactored later into a specific channel via a channel-master writer.
4-layer model:
/book-writer. Channel-agnostic CATEGORY tags (CLAIM / REFRAME / MECHANISM / EVIDENCE / STORY / APHORISM / PIVOT / OBJECTION). docs/extraction.mddocs/beat-method.md/authors-voice Apply Protocol with the operator's DEFAULT anchor (personal voice). Same machinery as every other writer; no anchor swap. The piece sounds like the operator./polish to 90/100, then /anti-ai, then a naive-reader pass (/congruence if installed, otherwise inline). Delegated, not duplicated./beat-writer only. No subcommands, no flag-syntax args. Internal dispatch from context./book-writer. AI never invents beats; the operator's head is the source./authors-voice./blog-writer). Workspace: [Project] Drafts (or reuse the operator's existing writing workspace if present)./blog-writer, /x-writer, /newsletter-writer, /book-writer). Refactor doesn't move the draft; the channel-master reads the Beats + Draft docs and re-shapes into its own container. See docs/refactor.md./polish → /anti-ai → naive-reader pass (/congruence if installed, otherwise read as a first-time reader and fix inline). Do not duplicate their logic.beat-writer/
├── SKILL.md (this file — router + firm rules + 4-layer model)
└── docs/ (loaded on routing match)
├── extraction.md (5-pass — channel-agnostic CATEGORY tags)
├── beat-method.md (beat = type + job + slot; commitments-not-content)
├── pipeline.md (extract → write → polish → anti-ai → congruence)
├── openwriter-surface.md (Drafts workspace + container + Beats/Draft doc pattern)
└── refactor.md (handoff to /blog-writer, /x-writer, /newsletter-writer, /book-writer)| User intent | Action |
|---|---|
| "/beat-writer" with no context | Ask ONE clarifying question: what are we writing about? |
| "draft something about X" / "write this" | Extract → shape → voice → polish |
| "I have this idea: [seed]" | Operator's dump = seed → 5-pass extraction → shape → write |
| "extract beats from [topic]" | Pass 1–5 only — no draft pour |
| "pour these beats: [list]" | Skip extraction; load beats; voice pour via /authors-voice |
| "polish this draft" | Skip extraction + write; run polish/anti-ai/congruence only |
| "make this into a blog/tweet/newsletter/page" | Refactor handoff — load draft + Beats, route to channel-master per docs/refactor.md |
{
"status": "draft-ready" | "needs-input" | "blocked",
"artifact": { "doc_id": "...", "workspace_id": "...", "container_id": "..." },
"next_steps": ["polish", "anti-ai", "refactor-to-<channel>", "publish-via-channel-master"],
"notes": "<optional>"
}/x-writer for tweet/article budgets./authors-voice owns Apply Protocol, NEVER-rule patching, fingerprint audit, blinder audit./authors-voice owns anchor setup, NEVER rules, fingerprints, corpus analysis./blog-writer, /x-writer, etc.)/authors-voice directly is faster/polish directly~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.