awesome-agv: Skill for Claude Code

Agent Protocols

Shared behavioral protocols for all agents in the workflow-team pipeline.

1. Recursive Nesting Protocol

When your scope card is too broad for a single context:

1. Further decompose using parallel-dispatch skill (§1 Decomposition, §5 Hierarchical Decomposition)
2. Spawn sub-agents with narrower scope cards using TypeName="self" with dedicated Role designations (see §3 Agent Spawn Protocol)
3. Your scope becomes the ceiling — children cannot operate outside it
4. Track sub-agent progress; merge results when all complete
5. Write .agentwork/handoff.md for your parent coordinator

Triggers for nesting:

• Task edits >3 unrelated files
• Scope card contains >2 features
• Context approaching 50% capacity
• Secondary expertise needed (delegate to specialist)

2. Pre-Implementation Restatement

Before writing code, restate in your own words:

1. What the .agentwork/brief.md / scope card asks you to build
2. What files you will create or modify
3. What assumptions you are making

If any assumption is uncertain, document it in your handoff and proceed with the conservative interpretation.

3. Agent Spawn Protocol (Universal TypeName="self" Pattern)

In the Antigravity subagent platform, TypeName="self" is the universal spawn mechanism that ensures subagents inherit their parent's full tool capabilities (file modification, command execution, and subagent dispatch for multi-tier nesting). Discovered named types default to read-only tool sets.

When spawning ANY agent type with a role file in .agents/agents/:

1. Always use TypeName="self" in invoke_subagent calls. This guarantees the subagent inherits write, execution, and subagent tools from the parent.
2. Differentiate roles via the Role field (e.g., Role: "Tech-Lead (Auth API)", Role: "Backend Engineer (Payments)", Role: "Test Automation Engineer").
3. Reference the role file in the system prompt — never paraphrase:

   "Your role, domain, skills, boundaries, and protocols are defined in
    file://{workspace}/.agents/agents/{agent-type}.md.
   Read this file FIRST before beginning any work."
   

4. The child agent MUST read the role file as its first action. Boundaries, domain rules, and workflow expectations are strictly defined by the role file.
5. Propagate this protocol recursively — Tech-Leads spawn specialized builders using TypeName="self" with builder role files, enabling deep parallel execution.
6. Subagent Monitoring & Reactivity: Parents receive automatic notifications when subagents send messages. Wait reactively or use manage_subagents to list active subagents. Do NOT use manage_task (which is exclusively for background commands/processes) for subagent lifecycle operations.