The Agent Skills Directory

[SAFE]: The skill is primarily composed of documentation, markdown templates, and basic shell script wrappers. The scripts provided (main.cjs, pre-execute.cjs, post-execute.cjs) contain boilerplate logic for environment detection and input validation with no suspicious side effects.
[DATA_EXFILTRATION]: While the skill interacts with the Sentry API (sentry.io), it does so using standard documented endpoints. It explicitly warns against hardcoding credentials and suggests using environment variables (SENTRY_AUTH_TOKEN), which is a security best practice.
[COMMAND_EXECUTION]: The SKILL.md file provides curl examples for interacting with Sentry. These are intended for the user to understand the API and do not represent arbitrary or hidden command execution by the agent itself.
[PROMPT_INJECTION]: No evidence of instructions attempting to bypass safety filters or override system-level constraints was found.

sentry-monitoring