The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local Ruby scripts to perform Sentry management tasks, such as listing projects and resolving issues.\n- [SAFE]: API tokens are stored locally at ~/.config/sentry/token with 0600 permissions, which is a secure practice for local credential management.\n- [DATA_EXFILTRATION]: Network activity is confined to the official Sentry API endpoints (sentry.io) for the purpose of retrieving and updating error monitoring data.\n- [SAFE]: The implementation relies exclusively on the Ruby standard library, minimizing the supply chain risk associated with external gems or packages.\n- [PROMPT_INJECTION]: The skill ingests data from external API responses (e.g., issue titles and event tags from Sentry) which could potentially contain malicious instructions.\n
Ingestion points: scripts/issues.rb, scripts/events.rb, and scripts/issue.rb fetch data from the Sentry API.\n
Boundary markers: Output is printed directly to stdout without delimiters or warnings.\n
Capability inventory: The skill executes shell commands for script invocation and performs network requests via Net::HTTP in scripts/auth.rb.\n
Sanitization: Data from API responses is printed to the console without sanitization or escaping.

sentry