Add learning_agents plugin implementation#217
Merged
Conversation
…ction chunks Introduces the learning_agents Claude Code plugin — auto-improving AI sub-agents that learn from their mistakes across sessions. This commit includes the architecture design doc, plugin manifest, instruction chunk docs (file structure, issue format, session logs, post-task reminder), and a prompt-review skill. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add the full LearningAgents plugin implementation: - Scripts: create_agent.sh for scaffolding new LearningAgents - Hooks: post_task.sh (tracks agent usage), session_stop.sh (suggests learning) - Agent: learning-agent-expert.md (meta-expert for learning cycle) - Skills: 7 skills implementing the learning cycle - User-facing: learning-agents (dispatch), create-agent, learn - Hidden: identify, report-issue, investigate-issues, incorporate-learnings - Update docs to use core-knowledge.md (replacing agent.yml), dashes throughout, and topic index listing instead of full content dumps Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Scaffold now creates additional_learning_guidelines/ with three empty files (issue_identification.md, issue_investigation.md, learning_from_issues.md) plus a README. Each file is auto-included via !`cat` in the corresponding hidden skill (identify, investigate-issues, incorporate-learnings), allowing per-agent customization of the learning cycle. Empty files produce no output. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
core-knowledge.md(plain markdown) instead ofagent.yml, standardizes on dashes throughout, and uses topic index listings instead of full content dumps in agent promptsWhat's included
create_agent.shhooks.json,post_task.sh,session_stop.shlearning-agent-expert.mdlearning-agents,create-agent,learnidentify,report-issue,investigate-issues,incorporate-learningsTest plan
create_agent.sh test-agentand verify scaffold output (directories, core-knowledge.md, agent file with topic index command)user-invocable: falseanddisable-model-invocation: true🤖 Generated with Claude Code