debug-sentry-monitor
Fail
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill directs the agent to search for and read sensitive files, specifically
.env,.env.local,.env.production, and.sentryclirc, for the purpose of extractingSENTRY_AUTH_TOKEN,SENTRY_DSN, and other authentication secrets. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources, creating a vulnerability surface:
- Ingestion points: Untrusted data enters the agent context via
sentry:get_sentry_resource,sentry:execute_sentry_tool(breadcrumbs and issue details),firecrawl:firecrawl_search, andfirecrawl:firecrawl_scrape. - Boundary markers: The instructions lack delimiters or explicit warnings to the agent to ignore instructions embedded within the Sentry logs or scraped web content.
- Capability inventory: The agent has capabilities to modify state via
sentry:update_issue, execute git commands, and run npm packages. - Sanitization: There is no mention of sanitization, validation, or escaping of external content before it is processed for root cause analysis.
- [COMMAND_EXECUTION]: The skill mentions the execution of
npx mushi-mushi, which downloads and runs code from a public package registry. It also utilizessentry:execute_sentry_toolto dynamically execute tools based on runtime discovery of names and argument schemas.
Recommendations
- AI detected serious security threats
Audit Metadata