The Agent Skills Directory

[SAFE]: A thorough security analysis of all files, including documentation and code, found no evidence of malicious behavior, obfuscation, or safety bypass attempts.
[COMMAND_EXECUTION]: The skill provides Python scripts (tam_sam_som_builder.py, survey_instrument_auditor.py, demand_signal_triangulator.py) for the user or agent to run locally. Analysis of these scripts confirms they use only standard Python libraries (such as argparse, json, and re) to process input data. They do not contain any dangerous sinks such as eval(), exec(), or subprocess calls, and they do not interact with the network or sensitive system files.
[DATA_EXFILTRATION]: There are no network operations or external data transmissions. All data processing occurs locally within the provided scripts using user-supplied JSON files.
[PROMPT_INJECTION]: The skill's instructions are purely operational and focus on market research tasks. No patterns associated with prompt injection, safety bypasses, or system prompt extraction were found.

market-research