Agent Sync OS
AgentSync OS is the operating system for autonomous agents. It provides a complete coordination layer so your AI agents can work together without stepping on each other. Core Primitives: Agent Identity — Register agents with unique IDs, API keys, and capability manifests. Discover agents by capability. Task Coordination — Priority queues with atomic claiming and a bidding system. Prevents…
2 subscribers
9.1/10 popularity
207 ms avg latency
95% success rate
15 endpoints
The in-depth APIMemo review for this API hasn't been published yet —
the data below comes straight from the public marketplace listing.
Agent Sync OS endpoints
| Method | Endpoint | Description |
|---|---|---|
| Agent Identity | ||
| POST |
/agent-register /agent-register |
Creates a new agent identity and returns an API key for subsequent requests. No `x-api-key` required — this is the bootstrap endpoint. |
| POST |
/agent-heartbeat /agent-heartbeat |
Updates the agent's `last_seen` timestamp to signal liveness. |
| GET |
/agent-discover /agent-discover |
Search for active agents, optionally filtered by capability. |
| Task Coordination | ||
| POST |
/task-claim /task-claim |
Claims a task by transitioning it from `pending` to `claimed` and assigning it to the calling agent. If the task is already claimed, a 409 conflict is returned and a conflict… |
| POST |
/task-bid /task-bid |
Submit a bid payload for a pending task. Multiple agents can bid before a task is claimed. |
| Resource Locking | ||
| POST |
/lock-acquire /lock-acquire |
Acquires a TTL-based lock on a resource. Expired locks are automatically cleaned up. Returns 409 if the resource is already locked by another agent. |
| POST |
/lock-release /lock-release |
Releases a lock on a resource. Only the owning agent can release it. |
| Shared Memory | ||
| POST |
/memory-write /memory-write |
Store contextual data scoped to an agent, task, project, or organization. |
| GET |
/memory-query /memory-query |
Retrieve memory entries, optionally filtered by scope and scope_id. |
| Messaging | ||
| POST |
/message-send /message-send |
Delivers a JSON payload to a recipient agent. The recipient must exist. |
| Governance | ||
| POST |
/reputation-update /reputation-update |
Adjusts the reputation score of a target agent by a delta amount. Score is clamped to 0–1000. An audit log and telemetry event are recorded. |
| POST |
/conflict-resolve /conflict-resolve |
Resolves a recorded conflict using a strategy: `priority` (highest reputation wins), `first_come`, or `random`. Audit log is created. |
| POST |
/policy-evaluate /policy-evaluate |
Checks all enabled policies (or a specific one) against the requesting agent and action. Supports rate_limit, access_control, execution, and resource policy types. |
| Simulation | ||
| POST |
/simulation-run /simulation-run |
Spins up simulated agents, tasks, locks, messages, and memory — then cleans up. Use to stress-test your coordination layer. Requires Authorization header (Bearer token). |
| Health | ||
| GET |
healthCheck /health |
Returns service status. No API key required. |
Agent Sync OS pricing
| Plan | Price | Rate limit | Quotas |
|---|---|---|---|
| BASIC | Free | — |
|
| PRO | $25 / month | — |
|
| ULTRA | $75 / month | — |
|
| MEGA | $150 / month | — |
|
