The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill processes untrusted external data which could potentially contain malicious instructions intended to influence the agent's reporting behavior.
Ingestion points: Data enters the agent context through Amplitude:search, Amplitude:get_experiments, and Amplitude:get_feedback_insights in SKILL.md.
Boundary markers: The instructions lack explicit delimiters or 'ignore embedded instructions' warnings for the data being interpolated into the report sections.
Capability inventory: The skill's capabilities are restricted to data retrieval and markdown report generation; it lacks file-write, shell execution, or unauthorized network capabilities.
Sanitization: There is no evidence of escaping or sanitization of external strings (e.g., experiment names or feedback themes) before they are included in the output.
[DATA_EXFILTRATION]: Data Access and Handling. The skill accesses experiment metadata and results within the context of the Amplitude platform.
No hardcoded credentials or unauthorized network operations were detected.
Data retrieval is performed using standard vendor-provided tools (Amplitude:*).

monitor-experiments