The Agent Skills Directory

[SAFE]: The skill implements secure credential management by using the SENTRY_AUTH_TOKEN environment variable and explicitly instructing users not to paste tokens into the chat interface.
[SAFE]: The bundled Python script scripts/sentry_api.py includes proactive data sanitization, automatically redacting PII such as email addresses and IP addresses from API responses before they are displayed to the user.
[SAFE]: All network operations are constrained to official Sentry API endpoints (sentry.io) using standard Python libraries, with no external dependencies or unverified third-party connections.
[DATA_EXFILTRATION]: Like any tool that processes external data, this skill has an attack surface for indirect prompt injection if Sentry event titles or messages contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Sentry API responses processed in scripts/sentry_api.py and returned to the agent.
Boundary markers: None identified in the prompt templates.
Capability inventory: Execution of the bundled Python script via subprocess.
Sanitization: PII redaction is present, but no specific sanitization for LLM instructions is implemented.

sentry