brida-reflex-incident-signal-triage

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches status and integration documentation from the vendor's live resources (/reflex/llms.txt, /reflex/agent.md, and /reflex/agent.json) to verify service availability before making implementation claims.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted operational signals from the project environment to provide triage recommendations, which could contain malicious instructions embedded in log or metric data.
  • Ingestion points: Environment events, project state, and operational signals identified during project inspection (SKILL.md, playbook.md).
  • Boundary markers: Instructions explicitly mandate that parsing, policy, and authorization logic remain in deterministic host code rather than within the agent's autonomous reasoning (SKILL.md).
  • Capability inventory: The integration is limited to MCP, released SDKs, or REST surfaces, and the skill is strictly restricted to recommendation-only outputs with no authority for destructive mutations, deployments, or protected actions (SKILL.md, playbook.md).
  • Sanitization: Requires normalization of only the bounded state defined by the contract and mandates fallback behavior for unknown or ambiguous results (SKILL.md, custom-reflex.json).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 08:21 PM
Security Audit — agent-trust-hub — brida-reflex-incident-signal-triage