The Agent Skills Directory

[Data Handling Guidance]: The skill includes a clear security note in SKILL.md advising the agent to treat all retrieved content from Sentry as untrusted data. This is an excellent safety measure to prevent indirect prompt injection from issue comments or error messages.
[Secure Pagination Handling]: The search script scripts/sentry_issues.sh implements a security check to ensure that pagination URLs (from the Link header) stay within the configured Sentry host. This prevents potential token exfiltration or redirection to malicious domains via manipulated API headers.
[Credential Management]: Authentication is handled by the platform runtime via a bearer token (SENTRY_TOKEN). The skill correctly instructs the user not to hardcode sensitive keys and uses a placeholder value for local testing, which aligns with secure credential management practices.
[Standard Tooling]: The skill relies on well-known system utilities (curl, jq) and a plain shell script to perform its operations, avoiding complex or unverifiable third-party dependencies.

sentry-api