experiment-design-planner

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for structured experiment planning and does not contain any malicious patterns such as prompt injection, obfuscation, or persistence mechanisms.
  • [EXTERNAL_DOWNLOADS]: The skill uses web search and fetch tools to gather information about current research leaderboards and benchmark conventions. These operations are used for information gathering and do not involve downloading executable content or scripts from untrusted sources.
  • [DATA_EXFILTRATION]: The skill writes experiment plans and metadata to the local project's docs/ and memory/ directories. This behavior is consistent with the skill's purpose as a research documentation tool and does not involve unauthorized transmission of data to external servers.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 12:40 PM