Meta-Analysis Methods
Guides the agent to combine and compare results across studies using effect sizes and confidence intervals.
- Repository
- aiming-lab/AutoResearchClaw
- Stars
- 14,802
- Licence
- MIT
- Files
- 1 file
What it does
When an agent is working on research or experiment results spread across several studies or runs, this skill makes it follow standard meta-analysis practice. It nudges the agent to report effect sizes rather than only p-values, use standardized metrics, account for differences in setups, datasets or random seeds, report confidence intervals, use forest plots, flag outliers, and consider publication bias. It is a short set of guidelines, not a calculator.
Great for
- Researchers comparing results from several experiments or papers
- Students learning how to aggregate findings across studies
- Anyone who wants an AI agent to present cross-experiment results more rigorously
Skip it if
- Simple single-experiment analysis, where these cross-study steps don't apply
- Situations needing verified statistical computation: the file contains only written guidance and no code, so the agent's math must still be checked by a person
- Medical or policy decisions based on the output, since the guidance is general and not a substitute for expert review
Try asking
- “Combine the accuracy results from these five experiments and report effect sizes with confidence intervals”
- “Make a forest plot comparing the cross-study results for these model variants”
- “Check whether my cross-experiment results are inconsistent and discuss possible outliers”
- “Summarise these runs across different random seeds and note heterogeneity”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
meta-analysis.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip meta-analysis.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip meta-analysis.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
- Give the agent the raw numbers or result files so it can compute effect sizes itself, and ask it to show the formulas used.
- State which metric you care about and how the studies differ (datasets, seeds, setups) so the heterogeneity discussion is useful.
- Verify any pooled estimates or confidence intervals independently before publishing, since the skill only sets expectations and does not run checks.
What's inside
- SKILL.md
Redistributed unmodified from aiming-lab/AutoResearchClaw 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.