By s-hiraoku
Complete plugin for Synapse A2A multi-agent framework including inter-agent communication, file safety, and history management
Design, review, and improve Claude/Codex skills based on Anthropic's "The Complete Guide to Building Skills for Claude". Use when creating a new skill, rewriting SKILL.md frontmatter and workflows, fixing under-triggering or over-triggering, designing scripts/references/assets, building test cases, or preparing a skill for upload/distribution.
Guide API design for REST, GraphQL, gRPC, and CLI interfaces. Use this skill when designing new APIs, reviewing existing API contracts, or establishing API conventions for a project. Produces consistent, well-documented API specifications.
Perform structured code reviews focusing on correctness, readability, security, and maintainability. Use this skill when reviewing pull requests, evaluating code changes, or establishing review standards for a team.
Simplifies and refines code for clarity, consistency, and maintainability while preserving all functionality. Focuses on recently modified code unless instructed otherwise. This skill should be used when code-quality checks pass but the code would benefit from structural cleanup — deduplication, branching simplification, naming improvements, or dead-code removal. Invoked as a subagent from /code-quality or directly via the Task tool.
Audit, restructure, and consolidate project documentation for clarity and maintainability. Use this skill when docs have grown organically and need reorganization, when duplicate content exists across files, or when documentation structure needs standardization.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
🌐 Language: English | 日本語 | 中文 | 한국어 | Español | Français
Enable agents to collaborate on tasks without changing their behavior
A framework that enables inter-agent collaboration via the Google A2A Protocol while keeping CLI agents (Claude Code, Codex, Gemini, OpenCode, GitHub Copilot CLI) exactly as they are
┌─────────────────────────────────────────────────────────────────┐
│ ✅ Non-Invasive: Don't change agent behavior │
│ ✅ Collaborative: Enable agents to work together │
│ ✅ Transparent: Maintain existing workflows │
└─────────────────────────────────────────────────────────────────┘
Synapse A2A transparently wraps each agent's input/output without modifying the agent itself. This means:
See Project Philosophy for details.
flowchart LR
subgraph Terminal1["Terminal 1"]
subgraph Agent1["synapse claude :8100"]
Server1["A2A Server"]
PTY1["PTY + Claude CLI"]
end
end
subgraph Terminal2["Terminal 2"]
subgraph Agent2["synapse codex :8120"]
Server2["A2A Server"]
PTY2["PTY + Codex CLI"]
end
end
subgraph External["External"]
ExtAgent["Google A2A Agent"]
end
Server1 <-->|"POST /tasks/send"| Server2
Server1 <-->|"A2A Protocol"| ExtAgent
Server2 <-->|"A2A Protocol"| ExtAgent
| Category | Feature |
|---|---|
| A2A Compliant | All communication uses Message/Part + Task format, Agent Card discovery |
| Agent Card Context Extension | Pass system context (ID, routing rules, other agents) via x-synapse-context to keep PTY clean |
| CLI Integration | Turn existing CLI tools into A2A agents without modification |
| synapse send | Send messages between agents via synapse send <agent> "message". Waits up to 30s for PROCESSING targets to finish before delivering (skipped for --force, priority 5, and --silent) |
| Sender Identification | Auto-identify sender via SYNAPSE_AGENT_ID env var → metadata.sender + PID matching (process ancestry, fallback) |
| Readiness Gate | /tasks/send returns 503 until agent initialization completes; priority 5 and replies bypass |
| Priority Interrupt | Priority 5 sends SIGINT before message (emergency stop) |
| Multi-Instance | Run multiple agents of the same type (automatic port assignment) |
| External Integration | Communicate with other Google A2A agents |
| File Safety | Prevent multi-agent conflicts with file locking and change tracking (visible in synapse list) |
| Agent Naming | Custom names and roles for easy identification (synapse send my-claude "hello") |
| Agent Summary | Persistent 120-char agent summary (synapse set-summary). Manual text, --auto from git context, or --clear. Visible in Canvas, MCP list_agents, Agent Card extensions.synapse, and synapse list --columns SUMMARY |
| Agent Monitor | Real-time status (READY/WAITING/PROCESSING/DONE), CURRENT task preview, terminal jump |
| Task History | Automatic task tracking with search, export, and statistics (enabled by default) |
| Quality Gates | Configurable hooks (on_idle, on_task_completed) that control status transitions |
npx claudepluginhub s-hiraoku/synapse-a2a --plugin synapse-a2aMulti-agent collaboration framework enabling Claude Code, Codex, and Gemini to communicate via Google A2A Protocol.
Survive context loss across sessions: SessionStart context injection, cost-ceiling, large-edit gate, session-end summary, ledger templates, and goal-manager workflow.
Severity-ranked code review and multi-pass security review with parallel specialist subagents.
Deterministic PreToolUse guards: secret detection, dangerous command blocking, branch protection, prompt-injection detection, and MCP tool allowlisting.
Verification feedback loop: format/typecheck/test on every edit, and block premature Stop until scripts/verify.sh passes.
Cross-agent messaging via SQLite. Send messages between CLI AI agents. No daemon, no network.
Enable AI agents to message, watch, and spawn each other across terminals with Claude Code, Gemini, and Codex.
Cross-harness LLM agent communication mesh — rooms, DMs, and presence
Marketplace for the Agency CLI - an AI agent orchestrator
Connect Claude Code to the Viche agent network — discover, message, and collaborate with other AI agents in real-time
Chain work across two independent agent processes that share no parent session. One agent passes a baton (a completion signal file) when done, optionally carrying a free-form payload; another waits for it, with a deadline, then starts its dependent task. Harness-agnostic protocol — each agent picks its own waiting mechanism. The baton tells you when, never what: the payload is content, not instructions. Provides /baton-pass and /baton-wait.