maa-pipeline-history-audit
Pass
Audited by Gen Agent Trust Hub on Nov 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses multiple Git commands (
git log,git show,git status,git diff-tree,git worktree) to inspect and manage the repository state during the audit process. - [DYNAMIC_EXECUTION]: In the validation phase (Step 6), the skill is instructed to execute a local Python script
tools\ci\check_resource.py. If the skill is used to audit an untrusted or malicious repository, this script could execute arbitrary code on the user's system. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the repository's Git history, JSON assets, and Python source files to generate its report.
- Ingestion points: Reads content from
assets/resource/**/*.json,agent/**/*.py,assets/interface.json, and output fromgit log. - Boundary markers: The instructions do not specify explicit delimiters or "ignore instructions" wrappers for the ingested content, though it is processed primarily via Python AST and JSON parsing.
- Capability inventory: The skill has the ability to execute shell commands (Git), run Python scripts (local validation tool), and write files to the
docs/directory. - Sanitization: The skill uses Python's
astmodule to parse code rather than executing it directly, which provides a layer of structural validation, though the validation step involves direct execution.
Audit Metadata