Analyze tournament debates and hypothesis reviews to produce actionable meta-reviews.
npx skills add https://github.com/achousal/EngramR --skill meta-review
Please help me install this Agent Skill. Skill: meta-review Source: https://github.com/achousal/EngramR/tree/main/.claude/skills/meta-review Command: npx skills add https://github.com/achousal/EngramR --skill meta-review
The Meta-Review agent analyzes tournament debates and hypothesis reviews to produce actionable meta-reviews that reveal patterns and guide improvement across the research cycle.
Invoke the meta-review agent on the latest tournament and hypothesis logs to generate a new meta-review.
High-intent search queries and answers about installing and using this skill.
You can extract recurring patterns from tournament logs by running a meta-review on files in _research/tournaments/ to identify winner patterns and recurring weaknesses, writing the structured output to _research/meta-reviews/{date}.md.
A meta-review is an analysis of tournament debates and hypothesis reviews that reveals recurring weaknesses and winner patterns, generating actionable recommendations to inform and guide future research iterations.
To generate recommendations from hypothesis review histories, parse the logs located in _research/hypotheses/ to synthesize recurring structural weaknesses into structured feedback for future /generate, /review, and /evolve cycles.
Yes, tournament match logs from _research/tournaments/ can be parsed and synthesized to produce meta-reviews that extract winner patterns and recommendations, directly informing future research cycle iterations.
The process leverages existing tooling located in the _code directory to parse, synthesize, and format meta-reviews, requiring no additional dependencies to transform raw logs into structured output.
You should run a meta-review when you have accumulated tournament debates and hypothesis logs and need to extract actionable meta-feedback to reveal patterns and guide improvements across the self-improvement loop.
dakesanGenerate testable scientific hypotheses with mechanistic explanations and experimental predictions.
LLM-Dev-OpsCoordinate multiple autonomous agents across research, development, and testing workflows.
GiorgioRicciardielloOrchestrate research goals, hypothesis generation, Elo debates, and meta-review feedback.
inbharataiGenerate qualitative interview guides with discussion outlines and thematic analysis frameworks.
