close-ticket
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, untrusted data from issue trackers and codebases to generate summaries and perform actions.
- Ingestion points: The agent is instructed to read the identity and scope of tickets, as well as relevant implementation evidence (
SKILL.md). - Boundary markers: No delimiters or specific instructions are provided to the agent to treat external ticket content as data only or to ignore embedded instructions within that content.
- Capability inventory: The skill has the capability to post comments and change ticket status using issue tracker tools or command-line interfaces (
SKILL.md). - Sanitization: There is no evidence of data sanitization or validation of the ticket content before it is processed by the agent.
- [PROMPT_INJECTION]: The instructions explicitly direct the agent to bypass standard interactive safety protocols.
- Evidence: The instruction "The request authorizes this comment and closure without another confirmation" (
SKILL.md) directs the agent to execute write operations (posting comments and closing tickets) without the standard human-in-the-loop verification usually required for autonomous actions.
Audit Metadata