The Agent Skills Directory

[SAFE]: The skill provides structured instructions for managing A/B tests through official PostHog tool calls such as experiment-create and experiment-update.
[SAFE]: No patterns of prompt injection, data exfiltration, obfuscation, or unauthorized remote code execution were detected.
[INDIRECT_PROMPT_INJECTION]: The skill represents a standard input surface for indirect prompt injection where untrusted data is processed.
Ingestion points: User-provided input for the experiment name, hypothesis (description), and feature flag key in SKILL.md.
Boundary markers: None explicitly defined for the interpolated user strings.
Capability inventory: Uses experiment-create, experiment-update, and experiment-launch as described in SKILL.md.
Sanitization: No specific sanitization or validation instructions are provided for the user-supplied metadata fields.

creating-experiments