Extract patterns and propose amendments to constitution.md and prompt.md after feature completion.
npx skills add https://github.com/ArvorCo/Relentless --skill learn-arvorco
Please help me install this Agent Skill. Skill: learn Source: https://github.com/ArvorCo/Relentless/tree/main/.claude/skills/learn Command: npx skills add https://github.com/ArvorCo/Relentless --skill learn-arvorco
This Skill captures and distills learnings from completed features to drive improvements in governance and prompting. It closes the loop after feature execution by formalizing lessons into actionable amendments.
High-intent search queries and answers about installing and using this skill.
To extract learnings from completed features, run the provided bash script on your feature directory to analyze progress.txt and prd.json, distilling patterns and failures into actionable amendment proposals for your governance files.
Turning feature run failures into constitution.md amendments involves analyzing completed feature data to identify error patterns, then generating formal proposals that append governance rules to prevent similar issues in future iterations.
Yes, you need jq installed to run the learnings extraction script, as it is a required dependency for parsing the structured feature data needed to generate governance and prompt proposals.
To apply approved proposals to prompt.md after feature completion, review the generated learning logs and manually propagate the accepted changes into your prompt.md file to refine future runs.
You should run the learnings extraction workflow immediately after feature completion, as it requires finished feature data to accurately capture costs and errors for iterative system improvement.
limaonOrchestrate autonomous feature development from prd.json with checks before commits.
auleewilliamsDraft or update project constitutions from repository context and user prompts.
zai-orgCapture AI session learnings, errors, and corrections into a .learnings repository.
sirkirbyEnforce OAK constitution amendment workflows and validation via the oak CLI.
