The Agent Skills Directory

[SAFE]: The skill implements legitimate mathematical models for revenue management, including unit price optimization, market segmentation, and EMSR capacity allocation. The code logic is transparent and aligns with the stated marketing optimization purpose.
[COMMAND_EXECUTION]: Legitimate use of the bash tool is identified for executing the provided Python scripts. These scripts perform localized mathematical calculations and do not invoke unauthorized subprocesses or shell commands.
[EXTERNAL_DOWNLOADS]: The skill depends on numpy and scipy. These are well-known, industry-standard scientific libraries for Python and are considered safe dependencies for optimization tasks.
[DATA_EXFILTRATION]: No unauthorized data exfiltration patterns were found. Network access is restricted to benchmark retrieval via standard web search tools, which is consistent with the skill's functional requirements.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from web searches and local files, creating a potential injection surface. However, this is mitigated by the following:
Ingestion points: Data enters via web_search and data/price_volume.csv (as referenced in the workflow).
Boundary markers: Logical boundaries are enforced by the separate script execution phase.
Capability inventory: Subprocess execution is limited to the provided calculation scripts; no arbitrary code execution or unsafe file writes are permitted.
Sanitization: The Python scripts (emsrb.py, markdown_linear_program.py, etc.) utilize strict type-casting (converting inputs to float) and produce structured JSON outputs, preventing malformed or malicious data from influencing the agent's logic or the dashboard's rendering.

price-demand-optimization