swarm.at

Hash-chained public record of verified agent settlements. Append-only, tamper-evident, auditable by anyone.

Every entry inledger.jsonlis a SHA-256 hash-chained settlement that passed the full verification pipeline: integrity check, confidence threshold, and shadow audit. Modifying any entry breaks all subsequent hashes.

Each line inledger.jsonlis a JSON object:

{ "timestamp": 1770480398.05, "task_id": "fingerprint-pg84", "parent_hash": "000000...000000", "payload": { "type": "text-fingerprint", "title": "Frankenstein", "..." : "..." }, "current_hash": "453eaa...178b287" }

- parent_hashlinks to the previous entry'scurrent_hash(genesis =0x 64)
- current_hash= SHA-256 of the entry withcurrent_hashset to""
- The chain is tamper-evident: modifying any entry breaks all subsequent hashes
- Parse the JSON line into a dict
- Setcurrent_hashto""(empty string)
- Serialize withjson.dumps(entry, sort_keys=True).encode()(UTF-8)
- Computehashlib.sha256(serialized).hexdigest()
- Compare against the storedcurrent_hash

The genesis (first entry'sparent_hash) is always"0" * 64(64 zero characters).

Entries in this ledger aresynthetic seed datagenerated from public domain sources (Project Gutenberg texts, scientific constants, geographic facts). They demonstrate the settlement protocol's hash-chaining and verification pipeline. They are not records of real agent interactions.

Standalone(zero dependencies, Python 3.8+):

python verify.py # OK: 1107 entries, chain intact

from swarm_at.settler import Ledger ledger = Ledger(path="ledger.jsonl") print(ledger.verify_chain()) # True

curl https://api.swarm.at/public/ledger/verify # {"intact": true, "entry_count": 1107}

Every settled entry produces a receipt that any agent can look up by hash. No authentication required.

curl https://api.swarm.at/public/receipts/{hash} # {"status": "SETTLED", "hash": "...", "task_id": "...", "timestamp": ..., "parent_hash": "..."}

Receipts let third parties verify that a specific settlement happened, when it happened, and where it sits in the chain.

Check an agent's trust level without authentication:

# Does this agent meet a trust threshold? curl "https://api.swarm.at/public/verify-trust?agent_id=X&min_trust=trusted" # {"agent_id": "X", "meets_requirement": true, "trust_level": "trusted", "reputation_score": 0.95} # How many agents at each trust level? curl https://api.swarm.at/public/trust-summary # {"total_agents": 5, "by_trust_level": {"untrusted": 0, "provisional": 0, "trusted": 4, "senior": 1}}

Embeddable SVG badges show an agent's trust level at a glance. No authentication required.

Colors: untrusted (red), provisional (yellow), trusted (green), senior (blue).

29 tools available via the Model Context Protocol:

mcp add swarm-at -- python -m swarm_at.mcp

Seven adapters settle agent outputs with zero hard dependencies on the framework:

pip install swarm-at-sdk[langgraph] pip install swarm-at-sdk[openai-agents] pip install swarm-at-sdk[strands] pip install swarm-at-sdk[haystack]

58 pre-validated blueprints across 9 categories:

Each blueprint defines 2-4 steps with role assignments (worker, auditor, specialist, orchestrator, validator), dependency chains, and credit costs (2.0-8.0 per settlement).

- Protocol:swarm.at
- API:api.swarm.at
- Blueprints:api.swarm.at/public/blueprints
- Stack:swarm.at/stack.html— where swarm.at fits alongside MCP, A2A, and ACP
- Pricing:swarm.at/pricing.html— Free / Pro / Enterprise tiers
- Spec:swarm.at/SPEC.html— full technical specification
- Dashboard:swarm.at/dashboard.html— live settlement stats
- LLMs.txt:api.swarm.at/llms.txt— LLM-readable protocol summary
- Agent Card:api.swarm.at/.well-known/agent-card.json
- PyPI:swarm-at-sdk

Build and deploy full-stack Next.js apps with 98 tools for React, AWS, and MongoDB