blueprint

Warn

Audited by Gen Agent Trust Hub on Aug 4, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The installation process involves cloning code from https://github.com/antbotlab/blueprint.git. This repository is not associated with a pre-verified trusted organization or the skill author's known vendor infrastructure.
  • [COMMAND_EXECUTION]: The skill provides a shell block intended for installation that executes commands including mktemp, git clone, and git checkout to retrieve external content.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes external data to generate construction plans.
    • Ingestion points: Scans the codebase and project memory (SKILL.md).
    • Boundary markers: Not explicitly defined in the provided instructions to prevent the agent from obeying instructions embedded in the codebase.
    • Capability inventory: Generates instructional context briefs for other agents and performs file system metadata checks (git ls-files, git status).
    • Sanitization: No explicit sanitization or filtering of ingested codebase content is described before interpolation into the generated plans.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 4, 2026, 01:17 PM
Security Audit — agent-trust-hub — blueprint