Indirect prompt injection into protected tool actions
Lower-trust retrieved content influences the agent into a tool operation outside the legitimate task or granted authority.
Candidate CWE: CWE-1427
Test hypothesis#
Compare a benign document with a copy containing a harmless stub-action instruction. Keep the owned victim task and tool set constant; compare attempted and committed mock events over a fixed small trial count.
Evidence required#
Exact legitimate task, edited source, document hash, acting identity, tool arguments and unauthorized synthetic side effect; report successes/attempts.
Rule out false positives#
Style changes, policy refusals and fabricated action claims do not prove a protected boundary was crossed.
Playbook and command references#
- AI tools: input trust, execution identity and approvals — Prove when lower-trust content changes a protected tool action, then locate the missing authorization or consent boundary.
Find the associated command snippets
Referenced by these target playbooks#
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.