mastermind-protocol

Mastermind Protocol

This file is a reference loaded by mastermind domain skills and master. It is NEVER invoked directly via the Skill tool.

---

Agent Delegation Protocol

Every Task/Agent spawn in mastermind and monomind MUST include the AGENT DELEGATION CAPABILITY block from mastermind-delegation/SKILL.md.

This makes delegation recursive: every spawned agent can itself spawn sub-agents, which can spawn their own sub-agents. The capability block tells each agent how to pick a real agent (the [PICK] line, mcp__monomind__pick, or local monomind pick) and how to delegate.

Placement: In the Task/Agent description or prompt field, insert the full == AGENT DELEGATION CAPABILITY == ... ================================= block immediately after BRAIN CONTEXT:.

Why this matters: Agents read their prompts cold. Without the delegation block, a spawned Security Engineer won't know it can delegate codebase exploration to a reviewer, or that it can spawn a Backend Architect to fix the issues it finds. With the block, every agent in the chain can self-organize and pull in the right specialist.

Reference: Full block text is in mastermind-delegation/SKILL.md. Copy it verbatim — do not paraphrase.

---

Brain Load Procedure

Execute at the START of every mastermind run (master or standalone domain command). Load in this order:

Step A — Tier 3 core principles (all domains): Try mcp__monomind__memory_hierarchical-recall with query "mastermind principles", topK 20. If it returns an error, fall back to: mcp__monomind__memory_pattern-search with query "mastermind principles", namespace "mastermind:principles", limit 20. Record the entry ids of whichever call actually returned results — results[].id from memory_hierarchical-recall, or patterns[].id from the memory_pattern-search fallback — into a running list: recalled_entry_ids.

Step B — Tier 2 weekly summary for this domain: Try mcp__monomind__memory_context-synthesize with query [current prompt keywords], maxEntries 10. If it fails, fall back to: mcp__monomind__memory_pattern-search with query [current prompt keywords], namespace "mastermind:<domain>:weekly", limit 10. memory_context-synthesize returns only a context string and a sources count — no entry ids — so it cannot add to recalled_entry_ids. Only the memory_pattern-search fallback contributes ids (patterns[].id); append them to recalled_entry_ids when that fallback runs.