The Agent Skills Directory

[SAFE]: The skill implements environment-aware error handling to prevent sensitive information disclosure.
Evidence: Instructions mandate hiding stack traces and internal error codes when the application environment is set to production.- [EXTERNAL_DOWNLOADS]: The skill refers to established and well-known error monitoring services and documentation.
Evidence: Mentions Sentry, Bugsnag, and Rollbar for production error tracking, which are reputable industry-standard services used for their intended purpose.- [INDIRECT_PROMPT_INJECTION]: The skill outlines a pattern for handling errors derived from external API requests, which represents an ingestion surface for untrusted data.
Ingestion points: API route handlers and centralized error middleware defined in src/middleware/error-handler.js.
Boundary markers: The skill suggests using RFC 7807 structured JSON responses to delimit error data.
Capability inventory: The skill uses Read, Write, Edit, Grep, Glob, and a restricted Bash tool.
Sanitization: Instructions in SKILL.md and implementation.md explicitly mandate input validation, request sanitization, and preventing raw stack trace exposure.

handling-api-errors