sentry
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external Sentry API endpoints which may contain untrusted content such as error messages or stack traces influenced by external actors.
- Ingestion points: The
scripts/sentry_api.pyscript fetches issue and event data from the Sentry API (e.g.,/api/0/projects/{org}/{project}/issues/). - Boundary markers: No specific delimiters or safety instructions are used to separate untrusted Sentry data from agent instructions.
- Capability inventory: The skill utilizes a bundled Python script for read-only network access to the Sentry API.
- Sanitization: The
redact_datafunction in the script filters for emails and IP addresses, but it does not sanitize the content for potential prompt injection patterns. - [COMMAND_EXECUTION]: The skill executes a bundled Python script
scripts/sentry_api.pyto perform deterministic API calls. This is the intended behavior for the skill's functionality. - [EXTERNAL_DOWNLOADS]: The Python script performs network requests to
sentry.io(or a user-configured base URL) to retrieve observability data. This connection to a well-known service is documented and necessary for the skill's purpose.
Audit Metadata