The Agent Skills Directory

[COMMAND_EXECUTION]: The skill specifies the execution of several local Python scripts (e.g., content_scorer.py, humanizer_scorer.py, brand_voice_analyzer.py) to process and analyze marketing materials. These scripts are invoked via the command line and take local files as input.
[INDIRECT_PROMPT_INJECTION]: There is an inherent surface for indirect prompt injection because the skill's tools are designed to ingest and process untrusted external data (markdown articles, ad copy JSON, and drafts).
Ingestion points: Files like article.md and ads.json are read by Python scripts in the marketing-skill/ directory.
Boundary markers: The provided documentation does not specify the use of delimiters or 'ignore' instructions for the processed data.
Capability inventory: The skill uses python3 to execute scripts that read and process file content.
Sanitization: No explicit sanitization or validation logic for the input files is described in the top-level SKILL.md.
[SAFE]: No malicious patterns such as credential exfiltration, remote code execution from untrusted sources, or obfuscation were detected. The skill's architecture relies on local files and a foundation context created by the user.

marketing-skills