agent-incident

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides an interface to incident.io, a well-known technology service. It follows security best practices for credential management by instructing the agent to use official CLI authentication or environment variables rather than hardcoding secrets.- [PROMPT_INJECTION]: The skill processes data from external incident.io sources (incidents, alerts, and service catalogs), which constitutes a potential surface for indirect prompt injection. However, the impact is minimized by strict command scoping that limits the agent's capabilities to the agent-incident binary.
  • Ingestion points: External data is ingested via commands like incident list, incident get, and alert list.
  • Boundary markers: No explicit boundary markers or isolation instructions are used for the external data.
  • Capability inventory: The agent is restricted to agent-incident subcommands and basic file reading/searching tools (Read, Grep, Glob).
  • Sanitization: The skill relies on the underlying CLI tool for data handling and does not perform additional sanitization of the API output.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 04:24 PM
Security Audit — agent-trust-hub — agent-incident