sentry-sdk-skill-creator

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute local shell commands for project environment detection (e.g., searching for package managers or framework configurations) and for verifying the integrity of generated research files (using wc, grep, and ls).
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the Sentry Wizard CLI via npx @sentry/wizard, which is an official tool from the vendor (getsentry) used to automate the integration of Sentry into various platforms. It also directs the agent to fetch documentation and source-level API details from official domains including 'sentry.io' and 'github.com/getsentry'.
  • [PROMPT_INJECTION]: The skill uses instructional markers like 'IMPORTANT' and 'CRITICAL' solely to guide the agent through its complex multi-phase research and creation process. No attempts to bypass safety filters or override system instructions were detected.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill specifically addresses the handling of sensitive environment variables like SENTRY_AUTH_TOKEN. It includes explicit instructions to add sensitive configuration files to .gitignore to prevent credential exposure.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 09:55 PM