AGENT SKILL

prisma-systematic-review

Use when conducting a systematic review or structured literature survey — protocol, PRISMA flow, screening log, and synthesis, wired to the project's literature tools. A systematic review is a *method* (predefined protocol, auditable screening, reproducible counts), not an ad-hoc reading pile — for casual exploration or a quick coverage pass, this skill is overkill.

License 许可
MIT
所属仓库
yibocat/prismnext
最近更新
2026-09-05
Stars
★ 4
Forks
0

01 前言 Front Matter

name
prisma-systematic-review
license
MIT
description
Use when conducting a systematic review or structured literature survey — protocol, PRISMA flow, screening log, and synthesis, wired to the project's literature tools. A systematic review is a *method* (predefined protocol, auditable screening, reproducible counts), not an ad-hoc reading pile — for casual exploration or a quick coverage pass, this skill is overkill.

02 资源文件 4

03 SKILL.md 正文

PRISMA Systematic Review

A systematic review is a method, not a reading pile: protocol first, auditable screening, reproducible counts. Where an ordinary review says "I read some papers and here is my take", a systematic review can answer "how did you search, what did you exclude, and why" — with logs. This skill wires the PRISMA 2020 discipline into the project's literature tools.

Not for: exploring a new area, building a reading list, or drafting an ordinary Related Work section (writing-related-work when enabled) — those need no protocol and no flow diagram.

Scaled-down mode: when full PRISMA is heavier than the occasion but coverage still needs to be defensible (e.g. a Related Work section someone may challenge), keep the two minimal records — the queries as executed and the exclusion reasons — and skip the protocol registration and flow diagram. Say that you are doing so.

When to use

  • The user wants a systematic review / systematic survey, not an ad-hoc one
  • A Related Work section needs defensible coverage ("how did you search?")
  • Meta-analysis preparation (screening + data extraction stage)

Files in this skill

  • references/prisma-2020-checklist.md — the 27-item checklist, condensed; read it when drafting the report, not before.
  • templates/protocol-outline.md — the protocol skeleton (write this FIRST).
  • templates/screening-log.csv — one row per record: source, query, id, decision, reason.
  • templates/prisma-flow.md — the flow-diagram counts file to keep current.

Workflow

  1. Protocol before searching — draft from templates/protocol-outline.md into the project (research question, PICO/SPIDER framing, inclusion & exclusion criteria, databases, date limits). Persist via research-brief-update or a protocol file after user confirmation.
  2. Search & log — run literature-discover per planned query; record each query and hit count in the screening log. Reproducibility means the queries are written down as executed.
  3. Screen in two passes — title/abstract pass then full-text pass; every exclusion gets a reason in the log. literature-read for abstracts; literature-read-pdf (intensive list) for full texts.
  4. Keep the flow counts current — identified → deduped → screened → excluded (reasons) → included. templates/prisma-flow.md is the file of record; the diagram is generated from it at report time.
  5. Include — papers that pass go to the library only with explicit user confirmation (literature-add); otherwise stage them for the session.
  6. Synthesize & report — draft the review following the checklist; citations via [@bibkey] / staged [n]; literature-export-bib for the manuscript.

Done when

  • Flow-diagram counts reconcile with the screening log (audited against each other).
  • Protocol deviations recorded with dates and reasons.
  • Included papers are in the library (user-confirmed) or staged; the bibliography is synced.
  • Scaled-down mode: the query log and exclusion reasons exist on disk — the rest is optional.

Rules

  • Counts in the flow diagram must reconcile with the screening log — audit them against each other before reporting.
  • Deviations from the protocol are allowed but must be recorded with dates and reasons.
  • Never silently re-run a query with different terms and keep the old count.

04 同仓库更多 Skills

yibocat
yibocat/prismnext

demo-multi-file

Local test skill with an extra notes file

★ 4
yibocat
yibocat/prismnext

writing-results

Use when drafting a Results / Experiments / Analysis section — narrating verified evidence from run receipts, with complete statistical reporting, derivation analysis, and honest treatment of negative results. Every number traces to a run; every claim survives the audit.

★ 4
yibocat
yibocat/prismnext

writing-related-work

Use when drafting a Related Work / literature review section or a standalone synthesis note — organizing prior work into a grounded, cited narrative or comparing methods across papers, from the project library plus staged external discovery. Never cites from memory.

★ 4
yibocat
yibocat/prismnext

writing-preliminaries

Use when drafting a Preliminaries / Background / Notation section — definitions, problem setup, and formal statements the rest of the paper builds on, with notation consistency and machine-checked formal claims. Reference patterns, not molds; include only what later sections actually use.

★ 4
yibocat
yibocat/prismnext

writing-methods

Use when drafting a Methods / Methodology / Approach / Model section — expounding the ideas and the reasoning behind the method, grounded in the project's brief, notes, and prior discussions, with empirical details from run receipts. Structure follows the idea's shape; reference angles, not fixed templates.

★ 4
yibocat
yibocat/prismnext

writing-introduction

Use when drafting or rewriting a paper's Introduction — framing the problem, positioning against prior work, and stating contributions as checkable claims, grounded in the brief, notes, and the library. Reference rhetorical patterns, adapted to the paper — never applied as molds.

★ 4