KS
Killer-Skills

opi — Categories.community

v1.0.0
GitHub

About this Skill

Perfect for Claude Code Agents needing automated learning extraction and improvement of future sessions. The most comprehensive Claude Code skills registry | Web Search: https://skills-registry-web.vercel.app

majiayu000 majiayu000
[0]
[0]
Updated: 2/20/2026

Quality Score

Top 5%
50
Excellent
Based on code quality & docs
Installation
SYS Universal Install (Auto-Detect)
Cursor IDE Windsurf IDE VS Code IDE
> npx killer-skills add majiayu000/claude-skill-registry/opi

Agent Capability Analysis

The opi MCP Server by majiayu000 is an open-source Categories.community integration for Claude and other AI agents, enabling seamless task automation and capability expansion.

Ideal Agent Persona

Perfect for Claude Code Agents needing automated learning extraction and improvement of future sessions.

Core Value

Empowers agents to automatically extract learnings from recent work using Git commands, such as analyzing fix commits and recent code changes, to improve future Claude sessions with enhanced workflow efficiency and accuracy through protocols like Git log and grep.

Capabilities Granted for opi MCP Server

Automating learning extraction from recent commits
Analyzing fix commits for improved code quality
Generating summaries of recent code changes

! Prerequisites & Limits

  • Requires Git access and configuration
  • Limited to analyzing recent commits and fix/refactor logs
Project
SKILL.md
3.3 KB
.cursorrules
1.2 KB
package.json
240 B
Ready
UTF-8

# Tags

[No tags]
SKILL.md
Readonly

Opi - Learning Extractor

Automatically extract learnings from recent work to improve future Claude sessions.

Workflow

Step 1: Gather Sources

Run these commands to collect learning material:

bash
1# Recent commits (last 20) 2git log --oneline -20 3 4# Today's commits with details 5git log --since="midnight" --format="%h %s" 6 7# Commits with "fix", "korjaus", "refactor" 8git log --oneline -50 | grep -iE "(fix|korja|refactor|bugfix)"

Step 2: Analyze Fix Commits

For each fix/refactor commit, examine what changed:

bash
1git show <commit-hash> --stat 2git show <commit-hash> -- "*.tsx" "*.ts" # Code changes

Look for patterns:

  • What was wrong? (the bug/issue)
  • What fixed it? (the solution)
  • Why? (root cause)

Step 3: Check Current Conversation

Review current conversation for:

  • User corrections ("ei, tee näin..." / "no, do it this way...")
  • Failed attempts that were fixed
  • Non-obvious solutions discovered
  • Repeated patterns

Step 4: Categorize Learnings

Determine where each learning belongs:

Learning TypeDestination
React/TypeScript pattern.claude/LEARNINGS.md
Supabase/DB pattern.claude/LEARNINGS.md
CI/CD specificci-doctor/references/learnings.md
Lint specificlint-fixer/references/learnings.md
Test specifictest-writer/references/learnings.md
Skill-specific[skill]/references/learnings.md

Step 5: Propose Learnings

Format each learning proposal:

markdown
1## Proposed Learning 2 3**Source:** commit abc123 / conversation 4**Category:** [Global / Skill-specific] 5**File:** [target file path] 6 7### [Issue Title] 8- **Pattern:** What triggers this 9- **Wrong:** ❌ The incorrect approach 10- **Right:** ✅ The correct approach 11- **Why:** Root cause

Step 6: Get Approval & Write

  1. Present all proposed learnings to user
  2. User approves/rejects each
  3. Write approved learnings to appropriate files
  4. Commit changes

Learning Quality Criteria

Only propose learnings that are:

  • Non-obvious - Claude wouldn't know this without being told
  • Actionable - Can be applied in future situations
  • Specific - Has concrete wrong/right examples
  • Recurring - Likely to come up again

Skip:

  • Typo fixes
  • One-off configuration issues
  • Project-specific constants
  • Things Claude already knows

Example Session

User: /opi

Claude: Let me analyze recent commits...

Found 3 potential learnings:

1. **useCallback for useEffect deps** (commit a1b2c3)
   - Source: AdminTranslationsPage infinite loop fix
   - File: .claude/LEARNINGS.md
   - Pattern: Functions in useEffect deps cause re-renders

2. **GitHub API dismiss format** (commit d4e5f6)
   - Source: CodeQL alert dismissal
   - File: ci-doctor/references/learnings.md
   - Pattern: Use "false positive" not "false_positive"

3. **Controlled Tabs state** (conversation)
   - Source: Feedback popup tab fix
   - File: .claude/LEARNINGS.md
   - Pattern: Radix Tabs need value+onValueChange for control

Add these learnings? [Y/n for each]

Commands Reference

bash
1# Commits since specific date 2git log --since="2025-01-10" --oneline 3 4# Commits by pattern 5git log --all --oneline | grep -i "fix" 6 7# Show specific commit 8git show <hash> --name-only 9 10# Diff for a commit 11git diff <hash>^..<hash>

Related Skills

Looking for an alternative to opi or building a Categories.community AI Agent? Explore these related open-source MCP Servers.

View All

widget-generator

Logo of f
f

widget-generator is an open-source AI agent skill for creating widget plugins that are injected into prompt feeds on prompts.chat. It supports two rendering modes: standard prompt widgets using default PromptCard styling and custom render widgets built as full React components.

149.6k
0
Design

chat-sdk

Logo of lobehub
lobehub

chat-sdk is a unified TypeScript SDK for building chat bots across multiple platforms, providing a single interface for deploying bot logic.

73.0k
0
Communication

zustand

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
Communication

data-fetching

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
Communication