posthog-security-basics
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFE
Full Analysis
- [CREDENTIALS_UNSAFE]: No hardcoded credentials were found. The skill uses clearly marked placeholders like
phc_abc123andphx_xyz789for illustrative purposes and correctly instructs users to manage secrets using environment variables and.gitignore. - [DATA_EXFILTRATION]: Network operations are limited to
curlcommands targeting the official PostHog API (app.posthog.com) for legitimate management tasks like creating keys and checking logs. No unauthorized data exfiltration patterns were detected. - [REMOTE_CODE_EXECUTION]: No remote code execution patterns or downloads of untrusted scripts were found. The skill provides static shell commands for the user to execute for configuration purposes.
- [PROMPT_INJECTION]: No prompt injection attempts, safety filter bypasses, or instructions to ignore system guidelines were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill serves as a documentation guide and does not ingest or process untrusted external data that could influence agent behavior.
- [OBFUSCATION]: No obfuscated code, hidden URLs, base64-encoded payloads, or homoglyph attacks were identified.
Audit Metadata