investigating-error-issue
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides SQL templates for the
posthog:execute-sqltool. These templates rely on the agent to insert user-provided strings likeissue_idandsession_id. If these inputs are not properly validated or sanitized by the underlying system, it creates a surface for SQL command injection. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and summarize external data from application exceptions and logs, which are potential ingestion points for malicious instructions.
- Ingestion points: The skill fetches data via
posthog:query-error-tracking-issue-eventsandposthog:query-logs. - Boundary markers: There are no instructions to use specific delimiters or protective warnings when the agent summarizes this external content.
- Capability inventory: The agent has access to multiple PostHog-specific tools, including
posthog:execute-sql. - Sanitization: The workflow does not include steps to sanitize or filter the content of error messages or logs before processing.
Audit Metadata