tailrocks-research

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external sources, including web pages, package registries, and third-party codebases.
  • Ingestion points: External URLs and Git repositories cloned during the research phase (SKILL.md, references/research-playbook.md).
  • Capability inventory: The skill can write files to the repository, execute Git commands (commit, push, clone), and use shell utilities like rg (SKILL.md, references/research-playbook.md).
  • Boundary markers: The instructions explicitly mandate that external content must be treated as evidence rather than instructions, and that embedded instructions must be flagged (SKILL.md, references/runtime-trust.md).
  • Sanitization: A multi-agent vetting process is used where independent citation-checkers in fresh contexts verify every claim against primary sources (references/research-playbook.md).
  • [EXTERNAL_DOWNLOADS]: The skill fetches external data by cloning repositories and accessing web content for research purposes.
  • Cloned projects are stored in disposable directories outside the main repository to prevent interference (SKILL.md, references/research-playbook.md).
  • [COMMAND_EXECUTION]: The skill uses Git for version control operations and shell tools for codebase analysis.
  • Actions like git commit, git push, and rg (ripgrep) are used according to a defined contract for delivering research artifacts (SKILL.md, references/research-playbook.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 03:12 PM
Security Audit — agent-trust-hub — tailrocks-research