macro-architecture

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a Python script (scripts/newadr.py) to automate the drafting of architectural decisions into markdown files. This script is distributed as part of the skill and operates locally on the filesystem using standard libraries.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze project source code, runtime evidence, and user requirements to provide architectural recommendations, which establishes an untrusted data ingestion surface.
  • Ingestion points: Repository source code, runtime logs/evidence, and user-provided constraints specified in SKILL.md and references/decomposition.md.
  • Boundary markers: The instructions do not define specific delimiters or instructions for the agent to ignore potentially malicious embedded content within the analyzed codebase.
  • Capability inventory: The agent can write markdown files to the local filesystem using the scripts/newadr.py utility.
  • Sanitization: Filenames are sanitized via slugification, but the generated ADR content does not undergo specific sanitization or filtering before being written to disk.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 12:27 AM
Security Audit — agent-trust-hub — macro-architecture