foundations-theory-of-constraints
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze user-provided system performance data, such as queue depths, utilization metrics, and 'Undesirable Effects' (UDEs), to diagnose organizational bottlenecks. While this creates a surface for indirect prompt injection, it is the primary purpose of the diagnostic tool.
- Ingestion points: Workflow instructions in
SKILL.mdand primitive templates (e.g.,01-five-focusing-steps.md) prompt the user to provide process data and UDEs. - Boundary markers: The skill does not implement explicit delimiters or 'ignore' instructions for the processed data, relying on the agent's internal reasoning.
- Capability inventory: The skill references a local Python script (
agents-skills-feedback-loop/scripts/append_learning.py) to append insights to alearnings.mdfile based on the analysis results. - Sanitization: No explicit sanitization or validation of the input data is defined within the instruction files.
- [EXTERNAL_DOWNLOADS]: The skill contains numerous references to external URLs for academic research and primary source material (e.g.,
arxiv.org,doi.org,google.com,itrevolution.com). All identified URLs are from well-known technology organizations or academic registries and are used for documentation purposes. - [COMMAND_EXECUTION]: The skill mentions the execution of a local utility script (
append_learning.py) as part of a standard feedback loop for agent self-improvement. This is a common platform pattern and does not involve downloading or executing remote code.
Audit Metadata