product-tracking-instrument-new-feature
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a documentation assistant, modifying only Markdown and YAML files within a dedicated
.telemetry/directory. No malicious patterns, such as remote code execution, unauthorized network access, or credential harvesting, were detected. - [PROMPT_INJECTION]: The skill instructions define a clear role and constraints for the AI agent. While the skill ingests external data (feature specs, PR descriptions), it uses this information solely to update documentation. It lacks any capability to execute instructions contained within that data as code or system-level commands.
- [DATA_EXFILTRATION]: The skill guidelines explicitly warn against including Personal Identifiable Information (PII) in event properties (found in
references/anti-patterns.md). It also limits the amount of raw data shared in the conversation to prevent inadvertent data leakage.
Audit Metadata