agentleFS
WilliamEbongWilliamEbong/CourseForge/.claude/skills/research-review/SKILL.md

research-review

Skill0 starsChanged 59 days agoMITView on GitHub

Reviewing research briefs and dossiers for coverage, source quality, currentness, jurisdiction and safety. Used by CourseForge RESEARCH_BRIEF and RESEARCH_DOSSIER reviewers.

18 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use resubmit-pipeline, experiment-bridge and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • raja21068Skill

    resubmit-pipeline

    Workflow 5: orchestrate a text-only resubmit of a polished paper to a different venue under hard constraints (no new experiments, no bib edits, no framework changes, never overwrite prior submissions). Phase 0 physical isolation, Phase 0.5 health + anonymity check, Phase 1 audit (proof / claim / citation), Phase 2 microedits via auto-loop with edit-whitelist + citation-audit --soft-only, Phase 3 kill-argument adversarial gate, Phase 4 final compile + Overleaf push via /overleaf-sync. Use when user says \"resubmit pipeline\", \"重投流程\", \"port paper to <new venue>\", \"resubmit to <venue>\", \"tighten paper for resubmission\", or has a rejected/withdrawn paper to move to a different top venue under tight time budget.

    Also in 5 projects that use this

  • raja21068Skill

    experiment-bridge

    Workflow 1.5: Bridge between idea discovery and auto review. Reads EXPERIMENT_PLAN.md, implements experiment code, deploys to GPU, collects initial results. Use when user says \"实现实验\", \"implement experiments\", \"bridge\", \"从计划到跑实验\", \"deploy the plan\", or has an experiment plan ready to execute.

    Also in 4 projects that use this

  • PostHogSkill

    experiment-audit

    Audit a PostHog A/B experiment for a customer — verify config, exposure, attribution, and metrics. Trigger phrases include \"audit [customer]'s experiment\", \"audit the [name] experiment\", \"check experiment setup for [customer]\", \"validate this A/B test\", or any request to review whether an experiment is correctly wired up. Assumes you already have MCP access to the customer's project (typically via the impersonation flow set up by the `impersonate-audit` wrapper that ships with this plugin).

    Also in 4 projects that use this

  • rongaredeSkill

    semantic-scholar

    使用 Semantic Scholar API 检索和验证学术论文。支持并发多关键词搜索、批量 ID/DOI 查询、引用分析、批量补全摘要、统一 Markdown 导出。覆盖 2.14 亿+ 学术论文,无需 API Key 即可使用。触发词:论文检索、论文验证、Semantic Scholar、S2 搜索、查论文、补全摘要、导出MD

    Also in 6 projects that use this

  • +8 more from raja21068

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 16 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every concept goal traces to a core question.

    line 38

“Common” is measured: kept by at least 6 in 10 of the 16 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Research review

How to review evidence efficiently

You cannot re-verify everything. Sample by consequence:

  1. Claims in category law-regulation and statistic.
  2. Claims marked high confidence that rest on a single source.
  3. Claims the course will obviously teach as headlines (definitions of core terms, key duties).
  4. Anything surprising, very round, or suspiciously neat.
  5. Sources with missing version, date or locator.

For each sampled claim: open the source, find the locator, and compare wording for scope, modality, numbers, population, date and jurisdiction.

Red flags

  • Fabrication signals: a URL that resolves to a different document or a generic landing page; a DOI that does not match the title; section numbers that do not exist; quotations you cannot find; publication dates that do not match the issuer's page.
  • Modality inflation: "should" becoming "must"; a framework or standard described as mandatory without the incorporating law.
  • Scope stretch: a rule for one sector, substance list, population or jurisdiction stated generally.
  • Qualifier loss: "where reasonably practicable", "listed substances", "in adults", "except" missing.
  • Laundered citations: a secondary source cited for a claim it attributes to a primary source that says something narrower.
  • Stale versions: repealed provisions, superseded editions, old statistics presented as current.
  • Synthesis as fact: inferences recorded as scientific or legal facts.
  • ID drift: the same document under two IDs, or one ID for two editions.

Coverage review

Map every brief question to the dossier text that answers it. "Mentioned" is not "answered": an answer states the fact, its mechanism or conditions, and its limits. Check the seams between sections, where parallel researchers each assume the other covered a topic. Check that every key term has a sourced definition, and that limitations and misconceptions are researched, not just the happy path.

Brief review

  • Every concept goal traces to a core question; every core question to a section.
  • The source hierarchy suits the subject and constrains weaker source types.
  • Currentness and jurisdiction requirements are explicit.
  • Safety boundaries are specific to the subject and name what must not be produced.

Writing good research findings

  • Location: the claim ID, source ID, section ID or question ID.
  • Evidence: quote the dossier text (short), then what the source actually says, with URL and locator.
  • Recommended action: the specific correction ("change 'must' to 'is recommended to' and recategorise as guidance-recommendation; cite AB-OHS-CODE s.7 only if the duty is statutory").
  • Severity: fabrication is a blocker; a wrong legal duty, wrong number or guidance-as-law is critical; missing qualifiers and weak sourcing on core points are major.

What validators already do

Code checks that every inline token resolves to a source record, that IDs are unique and well-formed, and that URLs are syntactically valid. Code cannot tell whether a source supports a claim, is current, or exists at that URL. That is the reviewer's job.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.