GHSA-j659-8xh6-5pq5
atomic-agents-stack: Parallel helper/delegate batch reserves $0 for models absent from the pricing table, bypassing the cost-cap fan-out guard
Impact
`_estimate_batch_cost` (`atomic_agents/agent.py`) looks up the per-model output price with `PRICING.get(model, {})`, returning 0.0 for any model not in the hardcoded pricing table. `_check_batch_reservation` then early-returns when the reservation is <= 0, skipping the batch reservation entirely. That reservation is the only defense against the documented fan-out race where every parallel helper/delegate reads the identical pre-batch on-disk cost total and each passes its individual check even though the collective spend overruns the configured cap. **Impact:** an operator running any model not in the pricing table (self-hosted/Ollama/vLLM, a new provider SKU) with `cost_guardrails` + `daily_cap_usd` set believes the cap protects them, but a single parallel batch can blow past the cap. The parallel-helper `model` argument can also be steered to an unknown id. The sibling `dream._estimate_dream_cost` does this correctly (`PRICING.get(model, _fallback_pricing())`), which makes this a clear defect. **Affected:** `agent.py` (`_estimate_batch_cost` / `_check_batch_reservation`), all versions through 1.0.0. **Fix:** use `PRICING.get(model, _costs._fallback_pricing())['output']` (mirror dream/calc_cost). Add a conformance test asserting an unknown-model batch reserves > 0 and that an over-cap unknown-model batch raises `CostGuardrailBlocked`.
An attacker can exploit this over the network, without logging in, without any action by a user. If it succeeds: high risk of data or systems being changed.
Affected packages
| Ecosystem | Package | Affected | Fixed in |
|---|---|---|---|
| pip | atomic-agents-stack | <= 1.0.0 | 1.1.0 |