The Agent Skills Directory

[SAFE]: The skill implements a robust dataset research workflow with explicit checks for data provenance, authority, and licensing. No malicious patterns, exfiltration attempts, or obfuscated content were detected.
[SAFE]: The helper script scripts/run_research.py provides deterministic logic for deriving workspace paths and extracting placeholders. It uses a restrictive regular expression ([a-zA-Z_][a-zA-Z0-9_]*) to identify placeholders, which inherently prevents arbitrary text or malicious instructions from being interpolated into the generated research brief.
[SAFE]: File system access in scripts/run_research.py is limited to reading the user-specified prompt file and deriving related directory structures within the project workspace. It does not perform any network operations or execute external code.

researcher-research