메뉴
Spawn PAI agents via MCP factory tool. Loads identity, injects RAG context, validates spawn chain, and executes via Task(). The bridge between MCP tools and Claude Code's agent spawning.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/majiayu000/claude-skill-registry --skill spawn-agent명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
SOC 직업 분류 기준
SKILL.md 표시 중
| name | spawn-agent |
| description | Spawn PAI agents via MCP factory tool. Loads identity, injects RAG context, validates spawn chain, and executes via Task(). The bridge between MCP tools and Claude Code's agent spawning. |
| model_tier | sonnet |
| parallel_hints | {"can_parallel_with":[],"must_serialize_with":["spawn-agent"],"preferred_batch_size":1} |
| context_hints | {"max_file_context":50,"compression_level":2,"requires_git_context":true,"requires_db_context":false} |
| escalation_triggers | [{"pattern":"spawn chain violation","reason":"Parent agent lacks authority to spawn requested child"},{"pattern":"identity.*not found","reason":"Agent identity card missing - needs creation"},{"keyword":["Deputy","opus"],"reason":"Deputy spawns require ORCHESTRATOR approval"}] |
Purpose: Spawn PAI agents using MCP factory pattern Created: 2026-01-16 Trigger:
/spawn-agent <agent_name> <mission>Model Tier: Sonnet (Execution)
This skill bridges MCP tools with Claude Code's Task() function for agent spawning.
The Pattern:
spawn_agent_tool prepares context (identity, RAG, skills)Task(prompt=spec.full_prompt, ...)Why This Exists:
/search-party instead)/agent-factory instead)/spawn-agent AGENT_NAME mission description here
/spawn-agent SCHEDULER Generate Block 10 schedule with ACGME compliance
/spawn-agent COMPLIANCE_AUDITOR Audit Block 10 for work hour violations
/spawn-agent G2_RECON Find all constraint implementations in the codebase
/spawn-agent COORD_ENGINE Optimize solver performance for schedule generation
/spawn-agent SCHEDULER Generate Block 10 --parent COORD_ENGINE
When /spawn-agent is invoked, follow these steps:
# First, load the spawn_agent_tool
MCPSearch(query="select:mcp__residency-scheduler__spawn_agent_tool")
spec = mcp__residency-scheduler__spawn_agent_tool(
agent_name="AGENT_NAME",
mission="The mission description",
context={"relevant": "context"}, # Optional
inject_rag=True, # Default: True
inject_skills=None, # Auto-match if None
parent_agent="PARENT_NAME" # Optional, for spawn chain
)
Check the returned spec for:
spawn_chain_valid: Must be True (or no parent specified)identity_found: Must be Truetier and model: Determine execution parametersIf validation fails:
if not spec["spawn_chain_valid"]:
# Escalate: parent cannot spawn this child
raise SpawnChainViolation(spec["spawn_chain_error"])
if not spec["identity_found"]:
# Escalate: need to create identity card first
raise IdentityNotFound(f"Create identity: .claude/Identities/{agent_name}.identity.md")
result = Task(
prompt=spec["full_prompt"],
subagent_type=spec["subagent_type"], # "general-purpose"
model=spec["model"], # haiku/sonnet/opus
max_turns=spec["max_turns"], # 5/20/50 based on tier
description=f"{spec['agent_name']}: {mission[:30]}..."
)
If the agent needs to persist state:
# Agent writes checkpoint to spec["checkpoint_path"]
# Example: .claude/Scratchpad/AGENT_SCHEDULER_20260116_143022.md
# To resume later:
# Read checkpoint, include in next spawn's context
| Tier | Model | Max Turns | Use Case |
|---|---|---|---|
| Specialist | haiku | 5 | Single-shot focused tasks |
| Coordinator | sonnet | 20 | Multi-step domain work |
| Deputy | opus | 50 | Strategic cross-domain work |
| G-Staff | sonnet | 15 | Advisory/research roles |
Rule: Match task complexity to tier. Don't spawn opus for simple validation.
The MCP tool validates that parent agents can spawn children:
# From agents.yaml
COORD_ENGINE:
can_spawn: [SCHEDULER, SWAP_MANAGER, OPTIMIZATION_SPECIALIST]
ARCHITECT:
can_spawn: [COORD_PLATFORM, COORD_QUALITY, COORD_ENGINE, ...]
If spawn chain fails:
| Spawner | Has Task()? | Spawn Method |
|---|---|---|
| ORCHESTRATOR (main session) | Yes | Use Task() directly |
| Deputy/Coordinator/Specialist (via Task) | No | Use CLI spawning |
| Agent spawned via CLI | Yes | Use Task() directly |
Subagents spawned via Task() do NOT have Task() available. They have all other tools (Bash, Edit, MCP, etc.) but cannot spawn further agents via Task().
Two working patterns:
Pattern A: Flat Parallelism (ORCHESTRATOR spawns all)
ORCHESTRATOR ──┬── Agent 1 (no Task(), works directly)
├── Agent 2 (no Task(), works directly)
└── Agent N (no Task(), works directly)
Pattern B: CLI Spawning (for hierarchical)
ORCHESTRATOR ── Coordinator (via Task, no Task())
│
└── Bash: claude -p → Specialist (HAS Task())
│
└── Can spawn further agents
When a subagent needs to spawn another agent:
claude -p "
You are SPECIALIST.
## Identity
[Include identity card content or key details]
## Mission
[Mission description with full context]
## Constraints
- Budget: Do not exceed this mission scope
- Report: Write findings to [path]
" --model haiku --max-budget-usd 1.00
Key points:
--max-budget-usd to prevent runaway costsBy default, inject_rag=True queries relevant context:
# spawn_agent_tool internally does:
rag_results = rag_search(
query=mission,
doc_types=agent.relevant_doc_types, # From agents.yaml
top_k=5
)
To disable (faster, less context):
spec = spawn_agent_tool(..., inject_rag=False)
Agents can persist state for resumption:
# .claude/Scratchpad/AGENT_SCHEDULER_20260116_143022.md
## Agent Checkpoint
**Agent:** SCHEDULER
**Mission:** Generate Block 10 schedule
**Status:** In Progress
**Timestamp:** 2026-01-16T14:30:22
### Progress
- [x] Loaded constraints
- [x] Ran solver (15 solutions found)
- [ ] Validation pending
### State
```json
{
"block_number": 10,
"solutions_found": 15,
"best_objective": 0.87
}
### Resuming from Checkpoint
```python
# Read checkpoint
checkpoint = Read(".claude/Scratchpad/AGENT_SCHEDULER_20260116_143022.md")
# Include in new spawn context
spec = spawn_agent_tool(
agent_name="SCHEDULER",
mission="RESUME: Complete Block 10 schedule generation",
context={"checkpoint": checkpoint}
)
Every spawn is logged to .claude/History/agent_invocations/:
{
"invocation_id": "20260116_143022_SCHEDULER",
"timestamp": "2026-01-16T14:30:22",
"agent_name": "SCHEDULER",
"tier": "Specialist",
"model": "haiku",
"mission": "Generate Block 10 schedule",
"parent_agent": "COORD_ENGINE",
"spawn_chain_valid": true,
"rag_injected": true,
"checkpoint_path": ".claude/Scratchpad/AGENT_SCHEDULER_20260116_143022.md"
}
Error: Identity card not found for agent: NEW_AGENT
Resolution:
1. Check if agent exists in .claude/agents.yaml
2. If not, use /agent-factory to create the agent
3. Create identity card at .claude/Identities/NEW_AGENT.identity.md
Error: Spawn chain violation: SCHEDULER cannot spawn ARCHITECT
Resolution:
1. SCHEDULER can only spawn: [] (no children)
2. ARCHITECT should be spawned by: ORCHESTRATOR
3. Escalate to correct parent or invoke as ORCHESTRATOR
Error: Agent registry not found at .claude/agents.yaml
Resolution:
1. Verify .claude/agents.yaml exists
2. Check for YAML syntax errors
3. Regenerate from identity cards if needed
/startupO and /startupO-lite can use this skill:
# In ORCHESTRATOR session
spec = spawn_agent_tool("G2_RECON", "Explore codebase for...")
Task(prompt=spec["full_prompt"], ...)
/coord-engine, /coord-platform, etc. spawn specialists:
# COORD_ENGINE spawning SCHEDULER
spec = spawn_agent_tool(
"SCHEDULER",
"Generate Block 10",
parent_agent="COORD_ENGINE" # Validates spawn chain
)
/search-party, /qa-party can spawn multiple agents in parallel:
# Spawn 10 G2_RECON probes
for i in range(10):
specs.append(spawn_agent_tool("G2_RECON", f"Probe {i}: {target}"))
# Execute in parallel
for spec in specs:
Task(prompt=spec["full_prompt"], ..., run_in_background=True)
| Tier | Agents (Examples) |
|---|---|
| Deputy | ARCHITECT, SYNTHESIZER |
| Coordinator | COORD_ENGINE, COORD_PLATFORM, COORD_FRONTEND |
| Specialist | SCHEDULER, COMPLIANCE_AUDITOR, TEST_WRITER |
| G-Staff | G1_PERSONNEL, G2_RECON, G3_OPERATIONS |
| SOF | SF_MEDIC, SF_ENGINEER, SF_WEAPONS |
# Scheduling
/spawn-agent SCHEDULER Generate Block X schedule
# Validation
/spawn-agent COMPLIANCE_AUDITOR Audit Block X for ACGME
# Research
/spawn-agent G2_RECON Find implementations of X
# Testing
/spawn-agent QA_TESTER Run test suite for X
# Documentation
/spawn-agent META_UPDATER Update docs for feature X
/spawn-agent (primary)/spawn (short form)/agent (alternative)spawn-agent: The bridge between MCP governance and Claude Code execution.