affaan-m/everything-claude-code

Metadata

• **Stars:** 152,433
• **License:** MIT
• **Last pushed:** 2026-04-12
• **Description:** The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

Archetype: mega-skill-pack

Structure

• skills/ — 181 skill directories spanning many domains
• agents/ — Agent definitions (38 agents per README)
• hooks/ — Hook implementations
• rules/ — Language-specific rule sets (common, typescript, python, golang, java, php, perl, kotlin, c++, rust)
• commands/ — 72 legacy command shims
• plugins/ — Plugin integrations
• contexts/ — Context files
• ecc2/ — Rust control-plane prototype (alpha)
• .claude-plugin/, .codex-plugin/, .cursor/, .gemini/, .opencode/ — Multi-harness support
• research/ — Research-first development
• mcp-configs/ — MCP server configurations
• schemas/ — Schema definitions

Extractable Value

Skills with Process Potential (cherry-pick, not bulk import)

From 181 skills, notable clusters that map to babysitter specializations:

**DevOps/Infrastructure:**

• docker-patterns, deployment-patterns, database-migrations, postgres-patterns

**Security:**

• security-scan (AgentShield integration), security-review, security-bounty-hunter, hipaa-compliance, healthcare-phi-compliance, defi-amm-security, llm-trading-agent-security

**Testing Methodologies:**

• tdd-workflow, e2e-testing, verification-loop, ai-regression-testing, eval-harness
• Language-specific testing: golang-testing, python-testing, rust-testing, cpp-testing, csharp-testing, kotlin-testing

**Domain Specializations:**

• Healthcare: healthcare-cdss-patterns, healthcare-emr-patterns, healthcare-eval-harness
• Finance: finance-billing-ops, evm-token-decimals, defi-amm-security
• Logistics: logistics-exception-management, returns-reverse-logistics, customs-trade-compliance, energy-procurement, inventory-demand-planning
• Legal/Compliance: hipaa-compliance, visa-doc-translate

**Agent/AI Meta-skills:**

• autonomous-loops, continuous-learning-v2, iterative-retrieval, context-budget, token-budget-advisor, cost-aware-llm-pipeline

**Content/Media:**

• manim-video, remotion-video-creation, frontend-slides, article-writing, content-engine

Methodological Patterns Worth Extracting

1. **Verification loop pattern** — Checkpoint vs continuous evals with pass@k metrics 2. **Continuous learning** — Auto-extract patterns from sessions into reusable skills 3. **Santa method** — Unknown, worth investigating 4. **Research-ops** — Research-first development workflow 5. **Hook profile system** — ECC_HOOK_PROFILE=minimal|standard|strict for runtime gating

Plugin Idea: Selective Skill Import

The selective install architecture (install-plan.js / install-apply.js) with manifest-driven pipeline and state tracking is a pattern worth studying for babysitter's plugin system.

What to SKIP

• ecc2/ Rust control-plane — Competing orchestration, not assimilable
• Memory/session infrastructure — SDK-covered primitives
• Cross-harness management — Covered by babysitter harness adapters
• Skill management/discovery — SDK-covered primitives
• Observer/dashboard — Babysitter has its own

Library Mapping

Plugin Marketplace Mapping