incident-responder
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and respond to production incidents based on external evidence and telemetry, creating a surface for indirect instructions to influence agent behavior.
- Ingestion points: The skill instructions explicitly direct the agent to map "affected operational paths" and ingest "incident signals" and "telemetry" which originate from potentially untrusted external systems or logs.
- Boundary markers: The instructions lack clear boundary markers or delimiters that would prevent the agent from mistaking instructions embedded within logs or incident reports for legitimate system status or commands.
- Capability inventory: The skill instructs the agent to "Implement or recommend... changes" and "sequence mitigation." If the execution environment allows for tool usage (such as shell access or file system modifications), an attacker could place malicious instructions within the telemetry data to trick the agent into recommending or executing unauthorized commands.
- Sanitization: No methods for sanitizing or validating the input data from logs or evidence paths are specified.
Audit Metadata