replication-package
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill identifies and executes a project's master script (e.g., 00_run_all.R, Makefile, or run.py) using Bash to verify reproducibility. This action is the primary purpose of the skill and is triggered by the user to validate research outputs.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from manuscripts and YAML metadata files to build replication documentation, creating an attack surface for embedded instructions.
- Ingestion points: Manuscript files ($0), YAML passport files (quality_reports/passports/*.yaml), and existing project scripts.
- Boundary markers: None identified for manuscript content interpolation.
- Capability inventory: File system read/write, Bash shell execution, and the ability to call other agent skills.
- Sanitization: No explicit sanitization or filtering of external content is described before interpolation into generated files.
Audit Metadata