Codify session trial and error into persistent agent instructions and enforcement mechanisms.
npx skills add https://github.com/kazukitcy/skills --skill codify-lessons
Please help me install this Agent Skill. Skill: codify-lessons Source: https://github.com/kazukitcy/skills/tree/main/codify-lessons Command: npx skills add https://github.com/kazukitcy/skills --skill codify-lessons
This skill solves the problem of repeating the same mistakes or re-deriving the same solutions across different sessions by codifying hard-won insights into machine-enforced rules.
Invoke the codify lessons skill to review the current conversation and propose improvements based on the trial and error encountered.
High-intent search queries and answers about installing and using this skill.
Codifying session failures into permanent agent rules requires pairing abandoned approaches with successful solutions, then routing those insights into standardized lint rules, runtime gates, or repository scripts for automated enforcement.
To stop repeating debugging mistakes across sessions, you can codify hard-won insights into machine-enforced rules, transitioning improvised debugging into persistent agent instructions and repository scripts.
Automating enforcement for workflow fixes involves routing codified lessons into the most effective tier, ranging from lint rules and runtime gates to repository scripts and instruction file updates.
Yes, codifying session lessons into lint rules requires runtime-specific reference files to validate destination locations and perform mandatory deduplication checks before applying updates.
Yes, you can capture API integration debugging fixes as instruction file updates or lint rules, ensuring future agents avoid the same trap immediately without re-deriving the solution.
Deduplication checks are mandatory when codifying agent instructions; failing to validate destination locations with reference files risks duplicating existing rules rather than creating new enforcement mechanisms.
mymx2Diagnose Kotlin and Spring failures by analyzing stack traces and startup logs.
79yuukiDesign AI agent harnesses with verification gates and feedback loops.
TienchingDisplay agent flow trace timelines and summaries for multi-agent debugging.
samuelmukotiAutomates MCP tool discovery and registration from external servers across stdio and HTTP transports.
