replication-package

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes system tools for directory synchronization, repository initialization, and document validation. Specifically, it employs rsync with symlink resolution, git for fresh history creation, and latexmk for compilation checks. All operations are scoped to a local output directory to ensure the original research project remains untouched.\n- [INDIRECT_PROMPT_INJECTION]: As the skill processes arbitrary user project files, it handles untrusted data that could potentially contain malicious prompts.\n
  • Ingestion points: Project files located at the path provided in the project-path argument (SKILL.md).\n
  • Boundary markers: The skill does not use specific delimiters for content analysis but relies on interactive user confirmation and dry-run reports before applying changes.\n
  • Capability inventory: The skill has access to shell utilities, file reading, and writing tools (SKILL.md) to scrub metadata and identity traces.\n
  • Sanitization: The skill identifies and removes sensitive infrastructure files (e.g., .env, .claude/) and scrubs specific attribution strings, though it does not sanitize the internal logic of the research scripts it processes.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 06:17 AM
Security Audit — agent-trust-hub — replication-package