ae-tracking-matrix
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from requirement documents (PRDs) in various formats as part of its 'Requirements Understanding' workflow (detailed in
references/workflows/prd-to-suggestions.md). This represents a surface for indirect prompt injection, where malicious instructions hidden in a PRD file could attempt to influence the agent's behavior. The skill mitigates this through strict functional constraints and mandatory confirmation steps before any data is modified. - [DYNAMIC_EXECUTION]: The orchestration script
scripts/prepare_delivery.pyusessubprocess.runto call an internal script,scripts/lint_tracking_plan.py, for automated validation. This use of subprocesses is confined to the local package environment and serves the primary purpose of auditing tracking plans. - [COMMAND_EXECUTION]: Several scripts in the package (e.g.,
scripts/finalize_plan.py,scripts/lint_tracking_plan.py) perform local file system operations and command-line checks to process Excel workbooks and maintain a local metadata index. These actions are transparent and aligned with the skill's documented capabilities.
Audit Metadata