debug-replay

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted interaction data and error messages ingested from external sources via session replays.
  • Ingestion points: Data enters the context through Amplitude:get_session_replay_events (interaction timelines) and Amplitude:get_session_replays (error messages/metadata) as described in SKILL.md.
  • Boundary markers: Absent. The instructions do not define delimiters or specific markers to isolate ingested external data from the agent's internal reasoning or instructions.
  • Capability inventory: The skill uses diagnostic and analytical tools (Amplitude:get_events, Amplitude:get_deployments, Amplitude:get_session_replay_events). No high-risk capabilities such as arbitrary command execution, file system writes, or unauthorized network operations were identified.
  • Sanitization: There are no explicit instructions for the agent to sanitize, escape, or validate the content of interaction timelines or error strings before processing them into reproduction steps.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 05:11 PM
Security Audit — agent-trust-hub — debug-replay