The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied topics and long-form articles to generate derivative content. This processing flow presents an attack surface where instructions embedded in the user-provided text could attempt to influence the agent's behavior during orchestration or quality review.
Ingestion points: User input provided as a seed topic or pasted long-form text for repurposing.
Boundary markers: Prompt templates in the prompts/ directory do not utilize explicit delimiters to isolate user-provided data from system instructions.
Capability inventory: The skill can perform literature searches via the PubMed client and perform file-writing operations to a designated local output directory.
Sanitization: The instructions do not describe explicit sanitization or filtering of input text before processing.
[EXTERNAL_DOWNLOADS]: The orchestrator.py script utilizes a PubMed client to fetch scientific literature metadata and abstracts from official sources. This is a core part of the skill's intended research functionality.
[COMMAND_EXECUTION]: The Python orchestration script performs local file system operations to create output directory structures and manage the content generation workflow. These actions are restricted to the local project environment and do not involve the execution of arbitrary external commands.

content-os