科研 Skill

研究项目结项清单

finishing-a-research-project

Runs the final completion checklist: rendering, bibliography, wiki, data, reproducibility, archiving. Use when a manuscript is finalized and ready to publish.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Finishing a Research Project

The project is done only when the last checklist is green. This skill walks the final mile: rendering, archiving, DOI, submission, and optional grant follow-up.

Announce at start: "Using finishing-a-research-project to close out <slug>."

For unmet items: tell the user which, ask for a short reason per skipped item, write reasons into knowledge/_meta/gate-overrides.log, and mark the project in the log as either status: done (with documented omissions) or status: paused. Hidden partial completion is not closure — explicit justification is.

Closing Checklist

  1. Manuscript renders cleanly — make render in output/<book|article>/ exits 0; PDF/HTML opens without warnings
  2. Wiki lint green — python scripts/lint-wiki.py exits 0
  3. BibTeX consistent — every citation key in the manuscript exists in references.bib; no unused keys (or unused are intentional)
  4. Peer review done — both constructive and adversarial reviewer reports archived under output/<target>/reviews/; all Major issues resolved or explicitly deferred with rationale
  5. Hypothesis addressed — pre-registered Hypothesis is explicitly confirmed, refuted, or marked "inconclusive — exploratory" in the manuscript
  6. Data reproducibility — output/data-analysis/ has requirements.txt or environment.yml; a README says "run python main.py to reproduce"; seeds are fixed
  7. Data archive — raw datasets in input/data/ are licensed; derivatives in output/data-analysis/results/ are documented
  8. Synthesis stable — all knowledge/synthesis/*.md referenced by the manuscript have status: stable
  9. AI-use disclosed — the manuscript (or its front matter / supplement) carries an AI-use statement in the form the target venue requires, drafted from the project's own records (see AI-Use Disclosure) and confirmed by the user
  10. Log closing entry in knowledge/_meta/log.md:
- YYYY-MM-DD · finish · <slug> · output=<path>
  1. Publication archive (DOI) — offer to deposit on Zenodo (or institution repository); update the log with DOI
  2. Decision: submission / follow-up — journal / publisher / conference? grant renewal?
  3. If follow-up grant: route to grant-finder
  4. If done done: commit everything (git add . && git commit -m "finish: <slug>"), tag the release

Process Flow

digraph finishing {
    "Render check" [shape=box];
    "Wiki lint" [shape=box];
    "BibTeX consistency" [shape=box];
    "Peer review archived" [shape=box];
    "Hypothesis resolution" [shape=box];
    "Data reproducibility" [shape=box];
    "Data archive" [shape=box];
    "Synthesis stable" [shape=box];
    "AI-use disclosed" [shape=box];
    "Any check fail?" [shape=diamond];
    "Fix and re-check" [shape=box];
    "Log closing entry" [shape=box];
    "DOI / archive publication" [shape=box];
    "Follow-up grant?" [shape=diamond];
    "Invoke grant-finder" [shape=box];
    "Commit & tag" [shape=box];
    "Done" [shape=doublecircle];

    "Render check" -> "Wiki lint";
    "Wiki lint" -> "BibTeX consistency";
    "BibTeX consistency" -> "Peer review archived";
    "Peer review archived" -> "Hypothesis resolution";
    "Hypothesis resolution" -> "Data reproducibility";
    "Data reproducibility" -> "Data archive";
    "Data archive" -> "Synthesis stable";
    "Synthesis stable" -> "AI-use disclosed";
    "AI-use disclosed" -> "Any check fail?";
    "Any check fail?" -> "Fix and re-check" [label="yes"];
    "Fix and re-check" -> "Render check";
    "Any check fail?" -> "Log closing entry" [label="no"];
    "Log closing entry" -> "DOI / archive publication";
    "DOI / archive publication" -> "Follow-up grant?";
    "Follow-up grant?" -> "Invoke grant-finder" [label="yes"];
    "Follow-up grant?" -> "Commit & tag" [label="no"];
    "Invoke grant-finder" -> "Commit & tag";
    "Commit & tag" -> "Done";
}

AI-Use Disclosure

Authorship stays with the humans: an AI system cannot be an author, because authorship presupposes accountability (DFG statement on generative models, 2023). What good scientific practice asks for instead is transparency proportionate to the use — which tools did which kinds of work, and that the authors checked the result and carry responsibility for it (Buck 2026, ch. 3.4). A full prompt log is not the standard: in a workflow where the model touches many steps, single prompts are no longer attributable to single passages, and the venue rarely asks for them.

Binding rules come first. Read the ## AI-Use Rules section of input/ideas/<slug>-design.md (recorded during brainstorming-research). If it is missing, ask the user once: which publisher / journal / university / funder rules apply, and which form do they require? Their rules override this template.

Draft from the record, not from memory. The project already documents what the model did:

  • knowledge/_meta/log.md — every ingest, synthesis, draft, review pass, with date
  • frontmatter author: human | llm | mixed on wiki pages, and based_on: on substitute sources
  • output/<target>/outline/*.md — which sections went through the drafting stages
  • output/<target>/reviews/ — the AI review passes and the user's decisions on them
  • knowledge/_meta/gate-overrides.log — where a gate was skipped (e.g. a straight-run draft)

Summarise by task, not by prompt: literature search, source acquisition, source extraction (ingest), synthesis, drafting, internal review, language editing. Name the model(s) and the host (Claude Code / OpenCode, provider). State what the humans did: chose the question and the thesis, read the sources the argument rests on, approved each argument and section, revised the prose.

Pick the form the venue wants — default to (a) if nothing is specified:

(a) Statement in the manuscript (acknowledgements, a footnote, or a section before the references):

## Use of AI tools

In preparing this <article / chapter / book>, <model, e.g. Claude Opus 5.5>
was used via <host> for: <literature search in <databases>>; <extracting
claims from sources into a project wiki, with page references>; <drafting
sections from an outline whose thesis and argument the author set and
approved section by section>; <internal review of the draft>. All sources
cited were checked against the original publications. The author(s)
reviewed and revised all AI-assisted content and take full responsibility
for the text.

(b) Tool overview page at the front of a thesis / dissertation: a table task → tool/model → purpose → how the output was checked.

(c) AI Usage Card (Wahle et al. 2023) as an appendix — a semi-standardised questionnaire; offer it for dissertations and where the institution asks for detailed documentation.

Present the draft, let the user correct it, and write the final text where the venue expects it. Do not cite the model as a source in the references — AI output is cited only where it is itself the object of study.

Reproducibility Statement Template

Drop into the manuscript (or supplementary material):

## Reproducibility

All analyses are reproducible from `output/data-analysis/`:

- **Code:** <files>
- **Data inputs:** `input/data/<files>` (licenses: <list>)
- **Environment:** Python <version>, see `output/data-analysis/requirements.txt`
- **Entry point:** `python output/data-analysis/main.py`
- **Random seed:** <fixed value>
- **Repository:** <git URL or Zenodo DOI>

Archive / DOI

Options:

  • Zenodo — free, accepts datasets + code + manuscripts; gives DOI immediately
  • Institutional repository — university OA repository
  • GitHub Release + Zenodo-GitHub integration — tag release, Zenodo creates DOI automatically

Pick one with user, update log entry with DOI.

Submission Handoff

If submission targeted:

  • Identify journal / publisher / conference CfP
  • Check author guidelines — may require format conversion (quarto render --to docx / --to latex, or pandoc)
  • Produce cover letter draft
  • Confirm corresponding author + ORCID
  • Do NOT submit on the user's behalf — prepare and hand over

Red Flags

ThoughtReality
"Wiki-lint is just a detail, I'll skip it"Broken wikilinks are the #1 reviewer feedback point.
"The hypothesis was wrong, I'll just reword it"Pre-registration principle: document honestly. The result is what it is.
"I'll do the Zenodo DOI later"Later = never. Now, or explicitly into the log.
"Follow-up grants are separate projects"Now is the best moment — the story is fresh in your head.
"It's just a working paper, not a publication"Even unpublished work must be reproducibly archived.
"AI use is obvious from the workflow, no need to state it"Readers and examiners cannot see the workflow. Undisclosed AI-generated text is a breach of good scientific practice, and most venues now require a statement.
"I'll attach the full prompt history to be safe"Not attributable in a multi-step workflow and rarely required. Disclose by task, in the form the venue asks for.

Key Principles

  • The checklist is the closure — not "feels done"
  • Hypothesis honesty — what was predicted, what came out, what's open
  • Reproducibility — manuscript AND data AND code AND environment
  • AI use is disclosed, authorship is not shared — proportionate, task-level transparency; the humans are the authors and carry the responsibility
  • DOI before submission — an archived version exists before the publisher process
  • Log entry is the final artefact — the project is closed in the log

示例提示词

我的手稿已通过评审,帮我跑一遍结项清单检查渲染、数据和归档。

所属插件

人文学科面向历史、哲学、语言学、考古与艺术研究:TEI/OCR 与数字档案、论证分析、语料库与语音学、放射性碳测年、乐谱与音频分析,以及人文论证写作与审稿。

安装该插件会一并获得同领域的 62 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

研究项目结项清单能做什么?

Runs the final completion checklist: rendering, bibliography, wiki, data, reproducibility, archiving. Use when a manuscript is finalized and ready to publish。

如何安装研究项目结项清单?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「研究项目结项清单」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

研究项目结项清单需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

研究项目结项清单经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。