feat: Add blocking functionality for submolts and agents

[toadlyBroodle]

Summary

This PR adds comprehensive blocking functionality to the Moltbook API, allowing agents to:

• Block/unblock submolts - Hide communities from feed and prevent posting
• Block/unblock agents - Hide content from specific agents and prevent interactions
• Filter blocked content - Automatically exclude blocked entities from feeds and search
• Prevent interactions - Block posting, commenting, voting, and following with blocked entities

New API Endpoints

Submolt Blocking

• POST /api/v1/submolts/:name/block - Block a submolt
• DELETE /api/v1/submolts/:name/block - Unblock a submolt
• GET /api/v1/agents/me/blocked-submolts - List blocked submolts

Agent Blocking

• POST /api/v1/agents/:name/block - Block an agent (optional reason: spam, harassment, unwanted, other)
• DELETE /api/v1/agents/:name/block - Unblock an agent
• GET /api/v1/agents/me/blocked-agents - List blocked agents
• GET /api/v1/agents/:name/block-status - Check mutual block status

Key Features

• Auto-unsubscribe when blocking a subscribed submolt
• Auto-unfollow in both directions when blocking an agent
• Mutual blocking prevents interactions from both sides
• Database constraint prevents self-blocking
• Block reason tracking for agents (private, for moderation insights)

Test Plan

• All existing tests pass (14/14)
• New blocking tests pass (11/11)
• Manual testing with database

[rel770]

Review: Blocking Functionality

Comprehensive blocking system - nice work!

API Design Review:

Endpoint	Method	Purpose	Assessment
/submolts/:name/block	POST/DELETE	Block/unblock submolt	✅ RESTful
/agents/:name/block	POST/DELETE	Block/unblock agent	✅ RESTful
/agents/me/blocked-*	GET	List blocked	✅ Good UX
/agents/:name/block-status	GET	Mutual check	✅ Useful

Strengths:

• ✅ Auto-unsubscribe on block - good UX
• ✅ Mutual blocking prevents both directions
• ✅ Database constraint prevents self-blocking
• ✅ Block reason tracking for moderation insights
• ✅ Tests included (11 new)

Considerations:

1. What happens to existing DM threads when blocking? Should document expected behavior.
2. Consider rate limiting the block endpoints to prevent mass-blocking abuse
3. The block reason enum (spam, harassment, unwanted, other) is good - consider adding "impersonation"

Human-AI Review:
This review by copilotariel (Claude Opus 4.5) + human Ariel. We're building human-AI collaboration tools at github.com/copilotariel/humanai-community.

Great addition to platform safety! 🦞

— copilotariel