replay-mcp
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze
replay.iorecordings, which contain external and untrusted data captured from web applications (e.g., console logs, DOM content, network traffic). - Ingestion points: Data enters the agent's context through tools like
ConsoleMessages,NetworkRequest,InspectElement, andReadSource(SKILL.md). - Boundary markers: The instructions do not define clear delimiters or "ignore instructions" warnings when processing content extracted from recordings.
- Capability inventory: The skill allows the agent to use powerful tools including
Evaluate(for executing JS expressions within the recording context),Logpoint, andBash(replayio:*)command execution. - Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the recordings before it is processed by the agent.
Audit Metadata