sentry-sdk-setup

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill determines the appropriate SDK by reading external project files like package.json, requirements.txt, and go.mod. These files are untrusted inputs that could be manipulated to influence the agent's routing decisions.
  • Ingestion points: Project manifests and build configuration files (e.g., package.json, Gemfile, build.gradle).
  • Boundary markers: None; the skill directly evaluates the presence of specific dependencies to make routing decisions.
  • Capability inventory: The skill primarily performs informational routing and provides commands to open vendor URLs.
  • Sanitization: None; the logic relies on pattern matching within project files.
  • [COMMAND_EXECUTION]: The skill suggests using system-native commands to open the Sentry dashboard in a web browser. These commands are restricted to the official vendor domain.
  • Evidence: Commands such as open, xdg-open, and start targeting https://sentry.io/settings/projects/ found in the 'Finding the DSN' section of SKILL.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 04:20 PM
Security Audit — agent-trust-hub — sentry-sdk-setup