Skill Learner
Skill Verified ActivePersistent correction system that captures mistakes from skills or Claude behavior and ensures they never repeat across sessions. Use whenever the user indicates something went wrong — "that's wrong", "fix this", "not what I wanted", "learn from this", "don't do that again", "esto está mal", "aprende de esto", "no hagas eso otra vez", or any correction, complaint, or teaching moment about Claude's output or a skill's behavior. Also use when the user asks to review, list, delete, or manage past corrections, or wants to create an improvement proposal for a skill author. Covers installed skills, general Claude behavior, and cross-session learning persistence.
To ensure that mistakes made by AI skills or general AI behavior are captured, understood, and prevented from recurring in future sessions, leading to a more reliable and personalized AI experience.
Features
- Captures and persists skill and general AI behavior mistakes
- Supports user corrections via specific feedback and trigger phrases
- Manages past corrections (review, delete, list)
- Optionally generates improvement proposals for skill authors
- Operates across sessions and conversational contexts
Use Cases
- When the user indicates something went wrong with a skill's output.
- When the user provides feedback on general Claude behavior that should be remembered.
- When the user wants to review or manage past corrections made to the AI's behavior.
- When the user wants to suggest improvements to a skill author based on observed mistakes.
Non-Goals
- Saving vague corrections without specific alternatives.
- Creating duplicate corrections for the same issue.
- Saving one-time preferences as universal rules without scope clarification.
- Exceeding a defined limit of active corrections to maintain performance.
- Saving corrections that bypass safety measures like input validation or authorization checks.
Versioning
- info:Release ManagementNo explicit versioning is declared in the SKILL.md frontmatter or manifest, and installation instructions reference main, but this is a common pattern for Claude skills.
Installation
npx skills add j4rk0r/claude-skillsRuns the Vercel skills CLI (skills.sh) via npx — needs Node.js locally and at least one installed skills-compatible agent (Claude Code, Cursor, Codex, …). Assumes the repo follows the agentskills.io format.
Quality Score
VerifiedTrust Signals
Similar Extensions
Janitor Report
100Full health check of all your skills in one report. Use when the user wants to check for errors, find duplicates, detect broken skills, or get a complete overview of skill health.
Janitor Audit
100Show all your installed skills
Hindsight Memory Skill
98Store user preferences, learnings from tasks, and procedure outcomes. Use to remember what works and recall context before new tasks. (user)
Investigate Capa Root Cause
100Investigate root causes and manage CAPAs (Corrective and Preventive Actions) for compliance deviations. Covers investigation method selection (5-Why, fishbone, fault tree), structured root cause analysis, corrective vs preventive action design, effectiveness verification, and trend analysis. Use when an audit finding requires a CAPA, when a deviation or incident occurs in a validated system, when a regulatory observation needs a formal response, when a data integrity anomaly requires investigation, or when recurring issues suggest a systemic root cause.
Soul Guardian
100Drift detection + baseline integrity guard for agent workspace files with automatic alerting support
Audit Dependency Versions
100Audit project dependencies for version staleness, security vulnerabilities, and compatibility issues. Covers lock file analysis, upgrade path planning, and breaking change assessment. Use before a release to ensure dependencies are current and secure, during periodic maintenance reviews, after receiving a security advisory, when upgrading to a new language version, before submitting to CRAN or npm, or when inheriting a project to assess its dependency health.