The Agent Skills Directory

[SAFE]: The skill references documentation and repositories owned by the verified vendor 'fingerprintjs', including docs.fingerprint.com and github.com/fingerprintjs repositories.
[CREDENTIALS_SAFE]: The skill correctly uses environment variables (FINGERPRINT_SECRET_API_KEY) for secret management and explicitly warns against logging secret keys or PII in event payloads.
[INDIRECT_PROMPT_INJECTION]: The skill provides logic to process external event data originating from the Fingerprint API to determine visitor trust levels.
Ingestion points: The evaluateSignals function in snippets/signals-policy.js accepts an event object that contains data from external API responses.
Boundary markers: Not explicitly required as the logic processes structured JSON fields (booleans, specific strings, and integers) rather than interpolating natural language data into prompts.
Capability inventory: No risky capabilities, such as file system writing, network requests to unknown domains, or command execution, are present in the provided logic.
Sanitization: The implementation uses direct field access and comparisons against a fixed set of expected signal values (e.g., checking if bot_info.identity is 'spoofed').

fingerprint-smart-signals