The Agent Skills Directory

[SAFE]: The skill follows professional strategy-consulting methodologies (PEST, SWOT, 4P) and provides strict guidelines for evidence handling, distinguishing between facts, inferences, and hypotheses. This disciplined approach reduces the risk of hallucinations or fabricated data.
[COMMAND_EXECUTION]: The skill utilizes local shell commands (soffice and pdftoppm) solely for the purpose of converting generated .pptx files to images for visual verification by the agent. This is a legitimate functional requirement for a presentation-generation skill.
[PROMPT_INJECTION]: While the skill processes untrusted external data (survey exports, scraped data), it includes robust mitigations such as mandatory source citations, specific evidence taxonomy requirements, and a dedicated 'Red team' review gate to verify claims against evidence.
[REMOTE_CODE_EXECUTION]: The skill utilizes a local Python library and utility scripts (scripts/check_deck.py, src/consulting_deck/deck.py) distributed with the skill. No external, unverified code downloads or remote script executions were found.

consulting-deck