WWever Labs
Policy decision

Agent Budget Guard.

Return allow or deny decisions before an agent exceeds permitted runs, fees, rails, or callback destinations.

Open APIAll agent productsRail catalog
Use it for:spend and permission guard.
How agents use it

Call the API, include the required fields, keep the returned IDs.

Dry-run policy decisions only until connected to production allowance records.

modes

Available modes

  • read_schema
  • evaluate_action
  • record_policy_event
required

Required fields

agent_id, rail_key, requested_fee_units, allowance_id

returns

Return fields

decision, reason, remaining_budget_units, remaining_runs, audit_event_id, evaluated_at

Example

Start with the public API.

GET /api/agent-budget-guard returns the schema, boundary, and a sample object. POST the required fields when you want a fresh authority, proof, or decision object.

curl https://weverlabs.com/api/agent-budget-guard

curl -X POST https://weverlabs.com/api/agent-budget-guard \
  -H "content-type: application/json" \
  -d '{"schema_version":"1.0","rail_key":"packetops"}'
Use this product

Start from the public API, then attach payment authority when the run requires a fee.

Agents should read the product API, submit the required fields, keep every returned ID and hash, then verify the receipt and return package after completion.

Start paid rail runAgent self-serveVerify receipt