By anouar1991
Automated learning pipeline that evaluates model sessions, extracts lessons, detects wasteful patterns, and optimizes CLAUDE.md rules
Review lessons for accuracy and completeness (Phase 2 only)
Evaluate a model session and extract lessons (Phase 1 only)
Full learning pipeline - evaluate session, extract lessons, optimize rules
Optimize and route lessons to rule files (Phase 3 only)
Reviews learning pipeline outputs for accuracy, completeness, and proper scope classification. Identifies missed opportunities and recommends corrections.
Analyzes agentic workflows and tool-use trajectories to identify execution errors and generate lessons learned for prompt optimization.
Analyzes learned lessons and optimizes CLAUDE.md rules following Claude best practices. Routes domain-specific rules to appropriate context files.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Automated learning pipeline that evaluates model sessions, extracts lessons, detects wasteful patterns, and optimizes CLAUDE.md rules following Anthropic best practices.
# In your project directory
claude
# Install the plugin
/plugin install rules-learning-pipeline
Or add to .claude/settings.json:
{
"plugins": ["rules-learning-pipeline"]
}
/rules-learning-pipeline:learn-from-session <session-file> [--dry-run] [--verbose]
Runs the complete 4-phase pipeline:
# Phase 1 only
/rules-learning-pipeline:evaluate-model <session-file>
# Phase 2 only
/rules-learning-pipeline:critics-review <lessons-file>
# Phase 3 only
/rules-learning-pipeline:optimize-prompts <lessons-file> [--dry-run]
| Agent | Purpose |
|---|---|
model-evaluator | Parse sessions, calculate metrics, extract lessons |
critics-reviewer | Validate scopes, check specificity, catch missed patterns |
prompt-optimizer | Route lessons, optimize rules, write files |
| Skill | Purpose |
|---|---|
wasteful-pattern-detection | Detect BDN, CIP, Verification Theater |
scope-classification | Classify lessons into correct scopes |
rule-optimization | Transform lessons into optimized rules |
| Pattern | Description |
|---|---|
| BDN | Broad check → dismiss → narrow check |
| CIP | Check → ignore errors → proceed |
| Verification Theater | "verify" → errors → "looks good" |
| Redundant Tool Chains | Multiple tools when one suffices |
| Scope Overkill | Full suite for single file change |
| Scope | Target | Applies To |
|---|---|---|
| user-general | ~/.claude/CLAUDE.md | Any project |
| cross-project-domain | ~/.claude/rules/*.md | Framework-specific |
| project | project/CLAUDE.md | This project |
| domain | context/rules/*.md | Project domain |
Automated learning pipeline that evaluates model sessions, extracts lessons, detects wasteful patterns, and optimizes CLAUDE.md rules
Comprehensive toolkit for Zimbra administration, mail flow management, API integration, and Zimlet development (Classic and Modern)
npx claudepluginhub anouar1991/rules-learning-pipelineSelf-learning system for Claude Code that captures corrections and updates CLAUDE.md automatically
Self-improving learning system that detects friction signals (corrections, repetitions, tool failures), extracts improvement candidates, and proposes rule/skill updates with explicit approval workflow.
Continuous learning hooks for gladiator MCP (observe patterns, reflect on them)
Session intelligence for Claude Code — find why your sessions fail and fix them
Analyze Claude Code agent session transcripts to identify inefficiencies, anti-patterns, repeated mistakes, missing tooling opportunities, and user frustration signals for continuous improvement
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses