The Agent Skills Directory

[SAFE]: The skill demonstrates secure secret management by instructing the use of environment variables (APIDECK_API_KEY, APIDECK_APP_ID) for authentication instead of hardcoding credentials.
[SAFE]: All external references and network endpoints (e.g., unify.apideck.com, bullhorn.github.io) target official vendor or service domains, with no evidence of typosquatting or malicious redirects.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from Bullhorn ATS, which serves as a potential surface for indirect instructions, although this is the intended primary purpose of the integration.
Ingestion points: External data enters the context via the apideck.ats.applicants.list method and the Proxy API mentioned in SKILL.md.
Boundary markers: The provided examples do not include explicit delimiters for the external data, relying on the agent's underlying safety filters.
Capability inventory: The skill uses the @apideck/unify SDK and curl for network requests and data retrieval from the ATS.
Sanitization: No specific sanitization or escaping logic is described in the integration examples.

bullhorn-ats