Agent skill

evidence-draft

Create per-subsection evidence packs (NO PROSE): claim candidates, concrete comparisons, evaluation protocol, limitations, plus citation-backed evidence snippets with provenance. **Trigger**: evidence draft, evidence pack, claim candidates, concrete comparisons, evidence snippets, provenance, 证据草稿, 证据包, 可引用事实. **Use when**: `outline/subsection_briefs.jsonl` exists and you want evidence-first section drafting where every paragraph can be backed by traceable citations/snippets. **Skip if**: `outline/evidence_drafts.jsonl` already exists and is refined (no placeholders; >=4 grounded comparisons per subsection in survey mode; `blocking_missing` empty). **Network**: none (richer evidence improves with abstracts/fulltext). **Guardrail**: NO PROSE; do not invent facts; only use citation keys that exist in `citations/ref.bib`.

View SKILL.md on GitHub Repository
Stars 377
Forks 25

Install this agent skill to your Project

npx add-skill https://github.com/WILLOSCAR/research-units-pipeline-skills/tree/main/.codex/skills/evidence-draft

SKILL.md

Evidence Draft

Build deterministic outline/evidence_drafts.jsonl packs from briefs + notes + optional evidence bindings.

Compatibility mode is active: this migration preserves the existing JSONL contract while moving evidence-quality policy, sparse-evidence routing, and evaluation-anchor rules into references/ and assets/.

Load Order

Always read:

  • references/overview.md
  • references/evidence_quality_policy.md

Read by task:

  • references/block_vs_downgrade.md when deciding whether thin evidence should block drafting or only downgrade claim strength
  • references/evaluation_anchor_rules.md when evaluation tokens, protocol context, or numeric claims are weak
  • references/examples_sparse_evidence.md for evidence-thin pack calibration
  • references/source_text_hygiene.md when paper self-narration or generic result wrappers are leaking into pack snippets / claim candidates

Machine-readable assets:

  • assets/evidence_pack_schema.json
  • assets/evidence_policy.json
  • assets/source_text_hygiene.json

Inputs

Required:

  • outline/subsection_briefs.jsonl
  • papers/paper_notes.jsonl
  • citations/ref.bib

Optional but recommended:

  • papers/evidence_bank.jsonl
  • outline/evidence_bindings.jsonl

Outputs

Keep the current output contract:

  • outline/evidence_drafts.jsonl
  • optional human-readable mirrors under outline/evidence_drafts/

Script Boundary

Use scripts/run.py only for:

  • deterministic joins across briefs / notes / evidence bank / bindings
  • snippet extraction and provenance assembly
  • policy-driven blocking_missing / downgrade_signals / verify_fields materialization
  • pack validation and Markdown mirror generation

Do not treat run.py as the place for:

  • filler bullets that make thin evidence look complete
  • hidden sparse-evidence judgment that is not inspectable from references/ / assets/
  • reader-facing narrative prose

Output Shape Rules

Keep these stable:

  • preserve the existing top-level pack fields already used by downstream survey pipelines
  • claim_candidates must remain snippet-derived
  • concrete_comparisons must remain genuinely two-sided; if one cluster has no usable highlight, drop the card and surface thin evidence upstream instead of fabricating an A-vs-B contrast
  • snippet sampling should stay cluster-aware: when a subsection has explicit clusters, evidence selection should avoid collapsing onto one route just because its abstracts contain louder result sentences
  • sparse evidence should surface as explicit blockers / downgrade signals / verify fields, not filler bullets
  • citation keys must remain constrained to citations/ref.bib

Compatibility Notes

Current mode is reference-first with deterministic compatibility:

  • assets/evidence_policy.json defines pack thresholds and sparse-evidence routing
  • assets/evidence_pack_schema.json documents/validates the stable pack shape
  • scripts/run.py still materializes the existing JSONL + Markdown outputs, but no longer pads sparse sections with generic caution prose

Quick Start

  • python .codex/skills/evidence-draft/scripts/run.py --workspace <workspace_dir>

Execution Notes

When running in compatibility mode, scripts/run.py currently reads:

  • outline/subsection_briefs.jsonl
  • papers/paper_notes.jsonl
  • citations/ref.bib
  • optionally papers/evidence_bank.jsonl and outline/evidence_bindings.jsonl
  • assets/evidence_policy.json and assets/evidence_pack_schema.json

Script

Quick Start

  • python .codex/skills/evidence-draft/scripts/run.py --workspace <workspace_dir>

All Options

  • --workspace <dir>
  • --unit-id <id>
  • --inputs <path1;path2>
  • --outputs <path1;path2>
  • --checkpoint <C*>

Examples

  • python .codex/skills/evidence-draft/scripts/run.py --workspace workspaces/<ws>

Troubleshooting

  • If packs look complete despite thin evidence, inspect assets/evidence_policy.json and references/block_vs_downgrade.md before changing Python.
  • If evaluation bullets are generic, inspect references/evaluation_anchor_rules.md and the policy asset.
  • If claims are strong but evidence is abstract/title-only, downgrade via downgrade_signals and verify_fields rather than adding narrative caveats.

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

WILLOSCAR/research-units-pipeline-skills

thesis-compile-review

对中文毕业论文进行编译、warning 分级、模板模式检查、数据与引用复查，并把问题回写成可继续迭代的 review checklist。 **Trigger**: 毕业论文编译检查, thesis compile review, warning 分级, 终稿复查, main.pdf 检查. **Use when**: 论文已经回写到 TeX 交付层，需要确认是否真正达到“可提交”的质量，而不是只做到能编译。 **Skip if**: 还处于中间层重构阶段，`chapters/*.tex` 尚未形成稳定交付稿。 **Network**: none. **Guardrail**: 不在这里重构章节主线；如果发现结构问题，明确回退到上游修复。

377 25
Explore
WILLOSCAR/research-units-pipeline-skills

front-matter-writer

Write the survey's front matter files (Abstract, Introduction, Related Work, Discussion, Conclusion) in paper voice, with high citation density and a single evidence-policy paragraph. **Trigger**: front matter writer, introduction writer, related work writer, abstract writer, discussion writer, conclusion writer, 引言, 相关工作, 摘要, 讨论, 结论. **Use when**: you are in C5 (prose allowed) and need the paper-like shell to stop the draft reading like stitched subsections. **Skip if**: `Approve C2` is missing in `DECISIONS.md`, or `citations/ref.bib` is missing. **Network**: none. **Guardrail**: no invented facts/citations; no pipeline jargon in final prose; no repeated evidence disclaimers; only use keys present in `citations/ref.bib`.

377 25
Explore
WILLOSCAR/research-units-pipeline-skills

thesis-question-list

维护中文毕业论文的 `codex_md/question_list.md`：把本轮问题、边界、优先级、协作方案和验收口径结构化，作为整条 thesis pipeline 的控制面。 **Trigger**: 毕业论文问题清单, thesis question list, 论文修改清单, 本轮目标, 结构问题梳理, review问题整理. **Use when**: 你已经有一批材料或上一轮 review 结果，需要明确这一轮到底修什么、不修什么，并给后续重构与编译复查提供统一入口。 **Skip if**: 当前只是在做一次性局部措辞修改，且没有形成新一轮结构/证据/编译问题。 **Network**: none. **Guardrail**: 不在这里写正文；不把问题单写成长篇散文；每条问题必须可执行、可验收。

377 25
Explore
WILLOSCAR/research-units-pipeline-skills

novelty-matrix

Create a novelty/prior-work matrix comparing the submission’s contributions against related work (overlaps vs deltas). **Trigger**: novelty matrix, prior-work matrix, overlap/delta, 相关工作对比, 新颖性矩阵. **Use when**: peer review 中评估 novelty/positioning，需要把贡献与相关工作逐项对齐并写出差异点证据。 **Skip if**: 缺少 claims（先跑 `claims-extractor`）或你不打算做新颖性定位分析。 **Network**: none (retrieval of additional related work is out-of-scope unless provided). **Guardrail**: 明确 overlap 与 delta；尽量给出可追溯证据来源（来自稿件/引用/作者陈述）。

377 25
Explore
WILLOSCAR/research-units-pipeline-skills

protocol-writer

Write a systematic review protocol into `output/PROTOCOL.md` (databases, queries, inclusion/exclusion, time window, extraction fields). **Trigger**: protocol, PRISMA, systematic review, inclusion/exclusion, 检索式, 纳入排除. **Use when**: systematic review pipeline 的起点（C1），需要先锁定 protocol 再开始 screening/extraction。 **Skip if**: 不是做 systematic review（或 protocol 已经锁定且不允许修改）。 **Network**: none. **Guardrail**: protocol 必须包含可执行的检索与筛选规则；需要 HUMAN 签字后才能进入 screening。

377 25
Explore
WILLOSCAR/research-units-pipeline-skills

rubric-writer

Write a rubric-based peer review report (`output/REVIEW.md`) using extracted claims and evidence gaps (novelty/soundness/clarity/impact). **Trigger**: rubric review, referee report, peer review write-up, 审稿报告, REVIEW.md. **Use when**: peer-review pipeline 的最后阶段（C3），已有 `output/CLAIMS.md` + `output/MISSING_EVIDENCE.md`（以及可选 novelty matrix）。 **Skip if**: 上游产物未就绪（claims/evidence gaps 缺失）或你不打算输出完整审稿报告。 **Network**: none. **Guardrail**: 给可执行建议（actionable feedback），并覆盖 novelty/soundness/clarity/impact；避免泛泛而谈。

377 25
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results