evidence-synthesis — Claude Code skill: what it does, how to install
Guides AI agents through structured rapid evidence reviews and literature assessments without fabricated citations.
npx skills add gasserane/personal-skills --skill evidence-synthesis -a claude-codeWhat evidence-synthesis does
This skill establishes a systematic protocol for rapid evidence assessments (REA) and literature syntheses in MEL and public health contexts. It begins by collecting core inputs: the precise research question, intended purpose, population boundaries, and timelines. Research questions are structured using PICO, SPIDER, or SPICE frameworks, while extracted findings receive GRADE-adjacent confidence tags from High down to Very low.
Inventing citations is strictly prohibited by the prompt instructions. The search routine tests for duplicate reviews via OpenAlex, evaluates vocabulary shifts across time periods, and weeds out retracted papers. Because OpenAlex cannot filter by study design or sample size directly, the workflow escalates complex methodological screening to Consensus. Without working database tools, the agent pauses to request documents from the user.
Conflicting study outcomes are deliberately preserved rather than averaged together. The output highlights opposing conclusions between papers and flags unresolved empirical debates using the ⚠️ Unresolved: marker.
What it helps with
- frame research queries using PICO, SPIDER, or SPICE frameworks
- conduct structured rapid evidence reviews for policy and evaluation design
- grade evidence certainty across findings using a GRADE-adjacent scheme
- map conflicting findings across scientific papers without smoothing differences
Example prompts for the agent
- Run a rapid evidence review on youth community health worker programs
- What does the evidence say about digital tracking tools for maternal care?
- Synthesize available evidence using a PICO frame for school-based health interventions
How to install evidence-synthesis
1Skills CLI (Vercel)
npx skills add gasserane/personal-skills --skill evidence-synthesis -a claude-code2Manually: put the files in place
~/.claude/skills/evidence-synthesis/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/evidence-synthesis/
1Skills CLI (Vercel)
npx skills add gasserane/personal-skills --skill evidence-synthesis -a codex2Manually: put the files in place
.agents/skills/evidence-synthesis/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add gasserane/personal-skills --skill evidence-synthesis -a cursor2Manually: put the files in place
.cursor/skills/evidence-synthesis/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install gasserane/personal-skills evidence-synthesis2Manually: put the files in place
.github/skills/evidence-synthesis/SKILL.md1Skills CLI (Vercel)
npx skills add gasserane/personal-skills --skill evidence-synthesis -a gemini-cli2Platform CLI
gemini skills install https://github.com/gasserane/personal-skills --path skills/evidence-synthesisManually: put the files in place
.devin/skills/evidence-synthesis/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add gasserane/personal-skills --skill evidence-synthesis -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about evidence-synthesis
How do I install evidence-synthesis in Claude Code?
Run npx skills add gasserane/personal-skills --skill evidence-synthesis -a claude-code in your terminal.
What does evidence-synthesis do?
It guides an agent through a rapid literature review process. The skill defines the search structure, extracts findings, grades confidence levels, and formats an analytical brief.
Can evidence-synthesis generate fake citations?
No. The skill rules strictly forbid fabricating sources, instructing the agent to pull from verified external searches or ask the user directly for literature files.
