a2a-agent-delegation
Skill0 starsChanged 6 months agoMITView on GitHub
Designs agent-to-agent delegation using A2A-style handoffs: task envelopes, capability cards, and async responses. Use when multiple agents collaborate, Google A2A, cross-framework agents, or delegating subtasks between services.
In this repository14 other files it gives its agents.
- agent-memory-patternsSkill
- agent-observabilitySkill
- agent-redteam-securitySkill
- browser-agent-automationSkill
- build-mcp-serverSkill
- create-production-skillSkill
- cursor-sdk-automationSkill
- deploy-agent-productionSkill
- evaluate-agent-qualitySkill
- github-ci-agentsSkill
- langgraph-multi-agentSkill
- multi-agent-supervisorSkill
- rag-agent-pipelineSkill
- secure-mcp-serverSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 15
Use both: MCP for databases/APIs.
- Distinctive: line 44
Do not chain >3 hops without human visibility
- Distinctive: line 60
Keep business logic in orchestrator graph.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
A2A Agent Delegation
Concept
MCP connects one agent to tools. A2A connects agent to agent for delegated work.
Use both: MCP for databases/APIs; A2A (or your orchestrator) for specialist agents.
Delegation envelope
Each handoff should include:
{
"task_id": "uuid",
"goal": "Summarize Q3 earnings call transcript",
"constraints": ["max 500 words", "cite timestamps"],
"context_refs": ["doc://transcript-123"],
"reply_to": "orchestrator/session-id",
"deadline_sec": 300
}Workflow
- [ ] Define agent capability card (what it accepts / rejects)
- [ ] Orchestrator validates task fits specialist
- [ ] Send envelope; receive task_id
- [ ] Poll or subscribe for completion (async)
- [ ] Merge result into orchestrator state; verify schemaOrchestrator rules
- Do not chain >3 hops without human visibility
- Idempotent task_ids for retries
- Propagate cancellation to child agents
- Log correlation IDs across agents
Failure modes
| Issue | Mitigation |
| Duplicate work | Idempotency keys |
| Stale context | Version context_refs |
| Runaway delegation | Max depth + budget cap |
Relation to LangGraph
- LangGraph supervisor node can call A2A client instead of inline subgraph
- Keep business logic in orchestrator graph; specialists stay replaceable
Pair with multi-agent-supervisor for in-process routing; A2A for cross-service.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.