The Agent Skills Directory

[SAFE]: No security issues detected. The skill implements standard API integration patterns and follows platform guidelines for secret management.
Authenticates securely using vm0_secrets and environment variables (ZENDESK_API_TOKEN).
Does not contain hardcoded credentials, using placeholders like your_api_token for documentation purposes.
[COMMAND_EXECUTION]: Employs common system utilities like curl, jq, and sed to perform legitimate API operations.
Uses curl for network requests and jq for processing JSON responses.
Uses sed to modify temporary request payloads in /tmp/zendesk_request.json before submission.
[EXTERNAL_DOWNLOADS]: Network communication is directed exclusively to official service endpoints.
All API interactions are targeted at *.zendesk.com, as documented in the skill instructions.
[SAFE]: Indirect prompt injection surface is restricted to processed ticket and user data.
Ingestion points: Support tickets, user profiles, and search results retrieved from the Zendesk API in SKILL.md.
Boundary markers: Absent, but operations are limited to standard support workflows.
Capability inventory: Network access via curl and file manipulation via sed and jq across all documented examples.
Sanitization: Content is processed as data via jq filters.

zendesk