Swarm At
Connect multi-agent products to swarm.at settlement, trust, and blueprint tooling over MCP with optional remote SSE or local PyPI stdio.
Overview
swarm-at is a MCP server for the Build phase that connects agents to swarm.at settlement, trust, blueprints, and swarm coordination tools.
What is this MCP server?
- Settlement protocol with hash-chained ledger and trust primitives for agent swarms
- Catalog lists 48 blueprints and 18 MCP tools for swarm coordination
- Remote SSE endpoint at https://api.swarm.at plus stdio via swarm-at-sdk (python -m swarm_at.mcp)
- Authenticated flows via SWARM_API_KEY for settlement, credits, and webhooks
- Registry description: 48 blueprints and 18 tools
- Server version 0.7.1; remote SSE at https://api.swarm.at
Community signal: 1 GitHub stars.
What problem does it solve?
Builders running agent swarms lack a standard MCP surface for settlements, credits, and auditable ledgers without bespoke API glue.
Who is it for?
Solo builders prototyping multi-agent products that need ledgered settlements, trust, and hosted or local MCP access.
Skip if: Single-agent chat wrappers with no credits, webhooks, or cross-agent accountability requirements.
What do I get? / Deliverables
After setup, agents can use swarm.at blueprints and tools with authenticated settlement and webhook-ready workflows.
- Access to 18 swarm.at MCP tools and 48 blueprints per registry description
- Configured local or remote MCP connection to swarm.at API
- Settlement- and credit-aware agent workflows with webhook hooks
Recommended MCP Servers
Journey fit
Agent swarm ledgers and settlement land in build when you orchestrate multiple agents, credits, and webhooks—not during initial idea research. agent-tooling is the shelf because the server is protocol infrastructure (ledger, trust, blueprints, tools) for running agent fleets.
How it compares
Agent-swarm settlement MCP platform, not a one-shot coding skill or a generic Git hosting integration.
Common Questions / FAQ
Who is swarm-at for?
Indie developers and small teams building multi-agent systems who want swarm.at ledger, trust, and blueprint tooling inside MCP clients.
When should I use swarm-at?
Use it during build when agents must settle work, track credits, or follow swarm blueprints against api.swarm.at.
How do I add swarm-at to my agent?
Install swarm-at-sdk, run python -m swarm_at.mcp for stdio or configure the SSE remote URL, and set SWARM_API_KEY for protected endpoints.