Enhance MCP development features and introduce skills management

- Added configuration options for requiring human approval before applying LLM-generated MCP patches.
- Updated Docker setup to include skills directory.
- Integrated skills management into the backend, allowing for procedural guides and skill matching.
- Refactored database initialization to apply Alembic migrations.
- Enhanced task approval process to handle MCP patch applications with optional approval.
- Introduced new schemas for skills and updated existing APIs to support skills functionality.

This commit lays the groundwork for improved agent capabilities and better management of MCP development processes.

.env.example

@@ -68,6 +68,8 @@ LLM_CLOUD_PROVIDER_ORDER=gemini,deepseek,openrouter
 
 # ---- MCP self-development (agent patches MCP source on failures) ------
 MCP_DEVELOPMENT_ENABLED=true
+# Require human approval before LLM-generated MCP source is applied/reloaded (recommended)
+MCP_DEV_REQUIRE_APPROVAL=true
 # Push to Gitea automatically after patch (false = approval required on task)
 MCP_DEV_AUTO_PUSH=false
 MCP_DEV_MODEL_TIER=premium
@@ -116,6 +118,11 @@ OBSIDIAN_SEARCH_PATHS=agentic-os-tasks,Meta/Agent-Memory
 # Troubleshooting decision rules (YAML)
 TROUBLESHOOTING_RULES_PATH=/app/rules/troubleshooting.yaml
 
+# Agent skills (procedural guides — skills/*/SKILL.md)
+SKILLS_PATH=/app/skills
+SKILLS_MAX_MATCHED=2
+SKILLS_MAX_BODY_CHARS=8000
+
 # ---- Frontend ---------------------------------------------------------
 # Public URL the browser uses to reach the backend API
 NEXT_PUBLIC_API_BASE_URL=http://localhost:8000