claim-evidence-matrix
Build a section-by-section claim–evidence matrix (`outline/claim_evidence_matrix.md`) from the outline and paper notes. **Trigger**: claim–evidence matrix, evidence mapping, 证据矩阵, 主张-证据对齐. **Use when**: 写 prose 之前需要把每个小节的可检验主张与证据来源显式化（outline + paper notes 已就绪）。 **Skip if**: 缺少 `outline/outline.yml` 或 `papers/paper_notes.jsonl`。 **Network**: none. **Guardrail**: bullets-only（NO PROSE）；每个 claim 至少 2 个证据来源（或显式说明例外）。
Installation
Details
Usage
After installing, this skill will be available to your AI coding assistant.
Verify installation:
npx agent-skills-cli listSkill Instructions
name: claim-evidence-matrix
description: |
Build a section-by-section claim–evidence matrix (outline/claim_evidence_matrix.md) from the outline and paper notes.
Trigger: claim–evidence matrix, evidence mapping, 证据矩阵, 主张-证据对齐.
Use when: 写 prose 之前需要把每个小节的可检验主张与证据来源显式化（outline + paper notes 已就绪）。
Skip if: 缺少 outline/outline.yml 或 papers/paper_notes.jsonl。
Network: none.
Guardrail: bullets-only（NO PROSE）；每个 claim 至少 2 个证据来源（或显式说明例外）。
Claim–Evidence Matrix
Make the survey’s claims explicit and auditable before writing prose.
This should stay bullets-only (NO PROSE). The goal is to make later writing easy and to prevent “template prose” from sneaking in.
Inputs
outline/outline.ymlpapers/paper_notes.jsonl- Optional:
outline/mapping.tsv
Output
outline/claim_evidence_matrix.md
Workflow (heuristic)
Uses: outline/outline.yml, outline/mapping.tsv.
- For each subsection, write 1–3 claims that are:
- specific (mechanism / assumption / empirical finding)
- falsifiable (“X reduces tool errors under Y evaluation”, not “X is important”)
- For each claim, list ≥2 evidence sources:
- prefer different styles of evidence (method paper + eval/benchmark paper, or two competing approaches)
- Keep it tight: claim → evidence → (optional) caveat/limitations.
- If evidence is weak or only abstract-level, say so explicitly (don’t overclaim).
- If
bibkeyexists inpapers/paper_notes.jsonl, include[@BibKey]next to evidence items to make later prose/LaTeX conversion smoother.
Quality checklist
- Every subsection has ≥1 claim.
- Each claim lists ≥2 evidence sources (or an explicit exception).
- Claims are not copy-pasted templates (avoid “围绕…总结…” boilerplate).
Helper script (optional)
Quick Start
python .codex/skills/claim-evidence-matrix/scripts/run.py --helppython .codex/skills/claim-evidence-matrix/scripts/run.py --workspace <workspace_dir>
All Options
- See
--help(this helper is intentionally minimal)
Examples
- Generate a first-pass matrix, then refine manually:
- Run the helper once, then refine
outline/claim_evidence_matrix.mdby tightening claims and adding caveats when evidence is abstract-level.
- Run the helper once, then refine
Notes
- The helper generates a baseline matrix (claims + evidence) and never overwrites non-placeholder work; in
pipeline.py --strictit will be blocked only if placeholder markers remain.
Troubleshooting
Issue: claims are generic or read like outline boilerplate
Fix:
- Tighten each claim to a falsifiable statement and add an explicit caveat if evidence is abstract-only.
Issue: you cannot add [@BibKey] because keys are missing
Fix:
- Run
citation-verifierto generatecitations/ref.bib, then use the produced keys in the matrix.
More by WILLOSCAR
View allPolish a single H3 unit file under `sections/` into survey-grade prose (de-template + contrast/eval/limitation), without changing citation keys. **Trigger**: subsection polisher, per-subsection polish, polish section file, 小节润色, 去模板, 结构化段落. **Use when**: `sections/S*.md` exists but reads rigid/template-y; you want to fix quality locally before `section-merger`. **Skip if**: subsection files are missing, evidence packs are incomplete, or `Approve C2` is not recorded. **Network**: none. **Guardrail**: do not invent facts/citations; do not add/remove citation keys; keep citations within the same H3; keep citations subsection-scoped.
Build per-H3 writer context packs (NO PROSE): merge briefs + evidence packs + anchor facts + allowed citations into a single deterministic JSONL, so drafting is less hollow and less brittle. **Trigger**: writer context pack, context pack, drafting pack, paragraph plan pack, 写作上下文包. **Use when**: `outline/subsection_briefs.jsonl` + `outline/evidence_drafts.jsonl` + `outline/anchor_sheet.jsonl` exist and you want to make C5 drafting easier/more consistent. **Skip if**: upstream evidence is missing or scaffolded (fix `paper-notes` / `evidence-binder` / `evidence-draft` / `anchor-sheet` first). **Network**: none. **Guardrail**: NO PROSE; do not invent facts/citations; only use citation keys present in `citations/ref.bib`.
Run this repo’s Units+Checkpoints research pipelines end-to-end (survey/brief/paper-review/evidence-review/idea/tutorial/graduate-paper), with workspaces + checkpoints. **Trigger**: run pipeline, kickoff, 继续执行, 自动跑, 写一篇, survey/brief/review/调研/教程/系统综述/审稿. **Use when**: 用户希望端到端跑流程（创建 `workspaces/<name>/`、生成/执行 `UNITS.csv`、遇到 HUMAN checkpoint 停下等待）。 **Skip if**: 用户明确要手工逐条执行（用 `unit-executor`），或你不应自动推进到 prose 阶段。 **Network**: depends on selected pipeline (arXiv/PDF/citation verification may need network; offline import supported where available). **Guardrail**: 必须尊重 checkpoints（无 Approve 不写 prose）；遇到 HUMAN 单元必须停下等待；禁止在 repo root 创建 workspace 工件。
Audit and rewrite evaluation/numeric claims to ensure they carry minimal protocol context (task + metric + constraint) and avoid underspecified model naming. **Trigger**: evaluation anchor checker, numeric claim hygiene, underspecified numbers, protocol context, 评测锚点检查, 数字断言, 指标上下文. **Use when**: before final merge/polish, or when reviewers would likely flag claims as underspecified (numbers without task/metric/budget), or `pipeline-auditor` warns about suspicious model naming. **Skip if**: evidence is too thin to justify numeric claims (route upstream to C3/C4), or you are pre-C2 (NO PROSE). **Network**: none. **Guardrail**: do not invent numbers; do not add/remove/move citation keys; if protocol context is missing, weaken/remove the numeric claim rather than guessing.