smithery/willoscar

writer-selfloop

Writing self-loop for surveys: run the strict section-quality gate, then rewrite only the failing `sections/*.md` files until the report is PASS.

Installation

$ npx skills add smithery/willoscar --skill writer-selfloop

Also in this package

Other skills from smithery/willoscar · top by installs.

npx skills add smithery/willoscar

Browse all from smithery/willoscar

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents codex

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 10,906 B
  • docs SUMMARY.md 763 B

History

  1. First recorded snapshot · 0 installs

SKILL.md

Writer Self-loop (fix only what fails)

Triggers & routing

  • Trigger: writer self-loop, writing loop, quality gate loop, rewrite failing sections, 自循环, 反复改到 PASS.
  • Use when: per-section files exist but C5 is FAIL/BLOCKED (thin sections, missing leads/front matter, citation-scope violations, generator voice).

Purpose: make writing converge without rewriting the whole paper.

This is the writing self-healing loop:

Default mode: semantic triage (LLM-first)

Treat the gate as a router, not as a reason to rewrite everything.

  • First, classify each failure as: evidence substrate vs writing execution vs voice contamination vs citation hygiene.
  • Then, fix the earliest responsible artifact (often not the merged draft).
  • Rewrite only the failing files; do not churn working sections.

If you do not run the helper script, you can still execute this skill by reading sections/*.md + packs and writing output/WRITERSELFLOOPTODO.md manually in the same PASS/FAIL style.

Role cards (use explicitly)

Writing Manager (triage + convergence)

Mission: make the draft converge by fixing only failing units and routing issues upstream when needed.

Do:

  • Rewrite only files flagged by the gate (one at a time).
  • If evidence is thin, route upstream instead of padding prose.
  • Keep citation keys and scope stable during rewrites.

Avoid:

  • Global rewrites that churn working sections.
  • Rewriting around evidence gaps with generic filler.

Router (earliest responsible artifact)

Mission: connect a visible writing failure to the earliest upstream fix.

Do:

  • Template voice -> local rewrite (subsection-polisher / front-matter-writer / chapter-lead-writer).
  • Thin evidence -> evidence-selfloop (notes/bindings/packs).

Avoid:

  • Repeatedly rewriting the same section when the pack cannot support it.

Role prompt: Writing Manager (triage + convergence)

You are the writing manager for a survey draft.

Your job is to make the draft converge by fixing only what fails, and by routing issues to the earliest responsible artifact.

Rules:
- If a section fails because the evidence is thin or out-of-scope pressure is high, do NOT pad prose.
  Route upstream to evidence-selfloop (notes/bindings/packs) and unblock the substrate.
- If a section fails because the prose is templated (narration openers, slide navigation, repeated disclaimer spam), rewrite locally.
- Never change citation keys during rewrites; keep scope local.

Working style:
- small, auditable edits
- one failing file at a time
- every fix either (a) improves argument moves, or (b) restores paper voice
  • the gate is deterministic (it inspects sections/ + sections/sections_manifest.jsonl)
  • the fixes are semantic (rewrite prose to execute argument moves and restore paper voice)

Inputs

  • sections/sections_manifest.jsonl (expected files + allowed citations)
  • sections/*.md (the actual prose units)
  • outline/writercontextpacks.jsonl (preferred per-H3 pack)
  • outline/evidence_bindings.jsonl
  • citations/ref.bib

Optional routing context:

  • output/EVIDENCESELFLOOPTODO.md (if evidence gaps are already known)
  • outline/subsectionbriefs.jsonl, outline/chapterbriefs.jsonl

Output

  • output/WRITERSELFLOOPTODO.md (report-class; always written)

Workflow

  1. Run the gate
  • Run the writer self-loop script (see Script section below). It reads sections/sections_manifest.jsonl and the actual sections/*.md files.
  1. Read the TODO report
  • Open output/WRITERSELFLOOPTODO.md and identify the failing file paths.
  • Treat repeated evidence-sized sentences within one H3 as blocking; repair support selection or remove the duplicate without changing citation scope.
  • For each failing H3, load its pack from outline/writercontextpacks.jsonl.
  • If a pack is missing or you need scope reminders, consult outline/subsectionbriefs.jsonl; for chapter-level context (leads/throughlines), consult outline/chapterbriefs.jsonl.
  1. Triage (writing vs evidence)

Route upstream (run evidence-selfloop) when:

  • output/EVIDENCESELFLOOPTODO.md already indicates missing anchors/comparisons for the same subsection
  • the pack in outline/writercontextpacks.jsonl is too thin to satisfy argument moves without guessing
  • you keep wanting out-of-scope citations outside outline/evidence_bindings.jsonl

Rewrite locally when:

  • the pack is rich enough, but prose is template-y (narration openers, slide navigation, repeated discourse stems)
  • the section has content but lacks a thesis, contrasts, evaluation anchoring, or limitations
  1. Use the right playbook for the failing file type
  • Front matter (sections/abstract.md, sections/S<sec_id>.md for Intro/Related Work, sections/discussion.md, sections/conclusion.md): use front-matter-writer.
  • Chapter leads (sections/S<secid>lead.md): use chapter-lead-writer.
  • H3 bodies (sections/S<sub_id>.md): use subsection-writer (draft) or subsection-polisher (local fix).
  1. Keep scope + citations stable
  • Keep citation keys unchanged and validate them against citations/ref.bib.
  • Stay in scope per outline/evidence_bindings.jsonl (prefer subsection-first).
  1. Rerun the gate until PASS

After PASS (merge-aware voice safety):

  • Proceed to style-harmonizer -> opener-variator ->

section-logic-polisher -> paragraph-curator -> evaluation-anchor-checker -> final argument-selfloop snapshot -> optional transition-weaver -> section-merger -> post-merge-voice-gate.

  • If post-merge-voice-gate FAILs with source: transitions, transition

insertion was explicitly enabled; fix outline/transitions.md and re-merge.

After PASS (mandatory style hygiene for all survey-family profiles):

  • Open output/WRITERSELFLOOPTODO.md and read ## Style Smells.
  • Run style-harmonizer on the listed sections/*.md files.
  • If it flags opener cadence / overview narration, then run

opener-variator on the listed files.

  • If it flags count-based limitation slots (e.g., Two limitations ...), run limitation-weaver on the listed files.
  • After logic and paragraph compaction, run evaluation-anchor-checker on the

settled H3 files, then regenerate the argument/manifest snapshot.

  • Keep meaning + citations fixed; treat this as surface-level rewrite, not new content.

How to fix a failing H3 (semantic recipe)

  • Use tension_statement + thesis from the pack to rewrite paragraph 1 (end with the thesis).
  • Use comparison_cards to write explicit A-vs-B contrast sentences.
  • Use evaluationanchorminimal / protocol snippets to add task/metric/constraint context.
  • Use limitation hooks to add a real caveat (not boilerplate).

Paper voice repairs (high-impact, low-risk)

Narration opener -> content claim:

  • Bad: This subsection surveys ...
  • Better: start with a tension/decision/lens sentence, then end paragraph 1 with the thesis.

Slide navigation -> argument bridge:

  • Bad: Next, we move from X to Y.
  • Better: Having established X, we can now examine how Y changes the trade-offs under comparable protocols.

Meta \"survey should\" -> literature-facing observation:

  • Bad: Therefore, survey comparisons should ...
  • Better: Across reported protocols, ... varies, which makes ... fragile unless ...

Disclaimer spam -> one policy paragraph + local caveat only when needed:

  • Keep evidence policy once in front matter; delete repeated \"abstract-only\" boilerplate inside H3.

Stop conditions (when rewriting is the wrong move)

Stop and route upstream if:

  • you cannot write a contrast or evaluation anchor without guessing
  • the pack lacks benchmarks/metrics/protocol details needed for the subsection\x27s core claim
  • fixing one H3 keeps forcing out-of-scope citations

Failure codes -> routing (how to read the gate)

The writer gate emits short issue codes. Treat them as routers and stop trying to pad prose around missing evidence.

First classify failures:

  • Evidence substrate: packs/anchors/comparisons are too thin to execute argument moves without guessing. Route upstream (evidence-selfloop).
  • Writing execution: packs are rich enough, but the section did not execute moves (thesis/contrast/eval/limitation). Rewrite locally.
  • Voice contamination: narration/slide/pipeline voice. Rewrite locally (opener + bridges), do not touch citations.
  • Citation hygiene: undefined keys / out-of-scope pressure. Fix citations/bindings or rewrite to in-scope; do not add new keys.

Quick map (common codes):

  • missingsectionsmanifest, emptysectionsmanifest, sectionsmissingfiles: run subsection-writer (create missing files + refresh manifest).
  • sectionsh3has_headings: remove headings; H3 bodies are body-only.
  • sectionsintro, sectionsrelatedwork_: rewrite front matter via front-matter-writer (dense positioning + one methodology note).
  • sectionsh3toofewparagraphs, sectionsh3too_short: if the pack is rich, expand by executing contrasts + eval anchor + limitation; if the pack is thin, route to evidence-selfloop.
  • sectionsh3missingcontrast: add explicit A-vs-B using comparisoncards; if cards are missing, route to evidence-selfloop (briefs/packs).
  • sectionsh3missingevalanchor, sectionsh3missingcitednumeric: add minimal protocol context (task/metric/constraint) in the same paragraph; if unknown, route upstream or weaken the claim.
  • sectionsh3missing_limitation: add a subsection-specific limitation; if none exists in the pack, route upstream.
  • sectionsh3narrationtemplateopener, sectionsh3slidenarration, sectionscontainspipelinevoice: rewrite openers/bridges to paper voice (no navigation commentary).
  • sectionsh3evidencepolicydisclaimer_spam: delete repeats; keep evidence policy once in front matter.
  • sectionscitesmissinginbib: rerun citation-verifier (do not invent keys).
  • sectionscitesoutside_mapping: rewrite to in-scope OR fix mapping/bindings (C2/C4) and rebuild packs; do not patch in the merged draft.
  • sectionsh3sparse_citations: if scope allows, plan adds via citation-diversifier then apply via citation-injector; if scope is too tight, expand mapping/bindings upstream.

Script (optional; deterministic gate)

Quick Start

  • uv run python .codex/skills/writer-selfloop/scripts/run.py --workspace <workspace>

All Options

  • --workspace <dir>
  • --unit-id <U###>
  • --inputs <semicolon-separated>
  • --outputs <semicolon-separated>
  • --checkpoint <C#>

Examples

  • Generate an actionable TODO list for failing sections/*.md:

- uv run python .codex/skills/writer-selfloop/scripts/run.py --workspace <workspace>