# Agent retry commits a protected single-use action twice

A response-loss or resume path duplicates an operation whose application contract requires at-most-once effect.

**Candidate CWE:** [CWE-841](https://cwe.mitre.org/data/definitions/841.html)

## Test hypothesis

In the local stub adapter commit one harmless operation, lose its response, then retry exactly once. Compare business operation IDs and committed events with the normal single-delivery baseline.

## Evidence required

Show the at-most-once contract, retry trigger, same authorized operation, persisted completion state and two distinct committed effects; distinguish request IDs from business identity.

## Rule out false positives

Intentionally repeatable calls or two independently authorized operations are expected. A duplicate displayed answer or repeated proposal without a committed side effect is insufficient.

## Playbook and command references

- [AI state: memory, resumed work, retries and private caches](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/modules/ai-state-lifecycle.html) — Test persistent context, task ownership, MCP request-state integrity, duplicate effects and caches across identities and time.
- [Workflow invariants, replay, idempotency, and races](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/modules/state-business-logic.html) — Find security consequences in workflow ordering and concurrency, with serial controls and durable-state evidence.

[Find the associated command snippets](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/commands.html?q=AI%20state%3A%20memory%2C%20resumed%20work%2C%20retries%20and%20private%20caches)

## Referenced by these target playbooks

[AI, agents & MCP](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/ai.html) · [Web & API](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/web-api.html) · [OSS & self-hosted](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/oss.html)

## Report checklist

- [ ] Capture the affected version, configuration, identities, and object ownership.
- [ ] Preserve a valid baseline and a negative control.
- [ ] Reproduce the forbidden data access, action, or state change on controlled fixtures.
- [ ] Explain the intended security rule and why the observed behavior violates it.
- [ ] State prerequisites, uncertainty, cleanup, and the smallest demonstrated impact.

## Conditional ATT&CK mapping

No direct ATT&CK mapping selected. Record the weakness and demonstrated impact without forcing an adversary-behavior classification. See the [mapping guide](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/attack.html).

## Primary references

- [MCP HTTP response loss and retries](https://modelcontextprotocol.io/specification/2026-07-28/basic/transports/streamable-http)

[Back to the complete vulnerability catalog](https://bug-bounty-field-manual.kmahshooq.chatgpt.site/vulnerabilities.html)