The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes curl to perform health checks against the Perplexity API and kubectl to manage Kubernetes secrets and roll out configuration changes during an incident.
[DATA_EXPOSURE]: The runbook includes debugging steps that output the first five characters and the total length of the PERPLEXITY_API_KEY environment variable to verify its status during auth failures.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the Perplexity API responses.
Ingestion points: API responses from https://api.perplexity.ai/chat/completions are captured in the triage and evidence collection steps in SKILL.md.
Boundary markers: None present; the responses are written directly to stdout or files.
Capability inventory: Uses curl, kubectl, dig, tar, and mkdir in SKILL.md.
Sanitization: No sanitization or validation is performed on the API responses before they are saved to files or displayed in the terminal.

perplexity-incident-runbook