The Agent Skills Directory

[SAFE]: The skill functions as a legitimate integration with the PopAI service. All network operations and file uploads target official vendor domains, and no suspicious behavior was identified.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user queries and external URLs. While this represents a standard attack surface, it is inherent to the functional purpose of the tool and is managed by the backend service.\n
Ingestion points: The SKILL.md instructions allow the agent to pass user-supplied topics, instructions, and URLs directly to the generate_ppt.py script via the --query argument.\n
Boundary markers: The skill does not implement specific delimiters or safety warnings for embedded instructions in user content.\n
Capability inventory: The skill uses python3 to execute a script that performs file reads (for uploads) and network POST requests to an external API.\n
Sanitization: The generate_ppt.py script does not perform sanitization on the query string before transmission to the PopAI API.\n- [EXTERNAL_DOWNLOADS]: The skill interacts with recognized vendor domains including api.popai.pro and popai-file.s3-accelerate.amazonaws.com for API calls and file storage.\n- [COMMAND_EXECUTION]: The skill executes the generate_ppt.py script using python3 to perform its primary tasks, which is standard for this type of agent skill.

popai-powerpoint-pptx