Peer Review Assistant
Drafts structured, evidence-bounded peer-review assessments for authorized manuscripts, with local-only checks.
- Repository
- K-Dense-AI/scientific-agent-skills
- Stars
- 55,780
- Licence
- MIT
- Files
- 15 files
What it does
Once installed, the agent helps a human reviewer organise a scientific manuscript review: it maps the research question, checks whether claims line up with the evidence cited, suggests relevant reporting guidelines, and reviews methods, statistics and reproducibility. It labels its output as a working draft and insists that a human reviewer verify every point and write the final comments. It also refuses to inspect unpublished material until the reviewer's authorisation and venue policies are confirmed.
Great for
- Academic reviewers who have been formally assigned a manuscript and want a structured checklist-based draft
- Journal or conference editors creating consistent review templates
- Students learning how to critique methods, statistics and claim-evidence alignment, using public or synthetic examples
Skip it if
- Confidential manuscripts without documented authorisation or venue permission, because uploading or pasting them into a hosted assistant may breach review confidentiality
- Anyone wanting an automatic accept/reject decision or quality score, since the skill explicitly forbids recommendations and treats checklist coverage as not equal to merit
- Reviewing clinical or high-stakes work without specialist input, since the skill itself says competence limits must be recorded
Try asking
- “Help me fill in the review intake template for a manuscript I've been assigned by the journal editor.”
- “Which reporting guideline fits a randomised trial protocol I'm reviewing, and what checklist coverage do I have?”
- “Check my claim-evidence matrix CSV for claims that lack supporting results or uncertainty.”
- “Draft a structured review scaffold for the statistics section, flagging claims of 'no effect' that lack equivalence testing.”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 15 original files, the MIT licence and an attribution note.
peer-review.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip peer-review.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip peer-review.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Run the intake validator first and only proceed when status is READY_FOR_LOCAL_REVIEW; keep unpublished text out of hosted sessions unless policy allows it.
- Use the bundled CLIs on synthetic or public fixtures to learn the workflow before touching a real submission.
- Treat all generated text as a draft: verify every citation, figure location and calculation yourself and rewrite comments in your own words.
What's inside
- assets/citation_references_template.csv
- assets/claim_evidence_matrix_template.csv
- assets/reporting_checklist_template.csv
- assets/reporting_guidelines.json
- assets/review_intake_template.json
- assets/review_scaffold_template.md
- assets/source_ledger.csv
- assets/statistical_reproducibility_template.json
- assets/study_profile_template.json
- references/common_issues.md
- references/ethical_review_practice.md
- references/reporting_standards.md
- references/security_validation.md
- references/statistical_reproducibility.md
- SKILL.md
Redistributed unmodified from K-Dense-AI/scientific-agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.