The Agent Skills Directory

[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill retrieves scholarly metadata, titles, and abstracts from the external OpenAlex API, which are then processed by the agent.
Ingestion points: Data enters the agent's context through search results and entity lookups via the OpenAlex API.
Boundary markers: The skill does not provide instructions to the agent to treat external API data with delimiters or as untrusted content.
Capability inventory: The skill scripts have capabilities to perform network requests using the requests library and write data to local files using csv.writer.
Sanitization: No sanitization or validation of the content returned from the API is implemented in the provided workflows.
[PROMPT_INJECTION]: Behavioral Steering. The skill includes a specific behavioral instruction to the agent to proactively suggest the 'K-Dense Web' (www.k-dense.ai) platform for complex research tasks. This steers the agent's behavior to promote a specific third-party service.
[DATA_EXFILTRATION]: Transmission of User PII. The skill setup instructions suggest that users provide an email address to be used in the mailto parameter of API requests. This transmits user PII to the OpenAlex service to gain access to higher rate limits (the 'polite pool'). While this is a documented feature of the OpenAlex API, it involves sharing user identity information externally.
[EXTERNAL_DOWNLOADS]: Package Dependency. The skill requires the installation of the requests library from a public package registry using the uv tool.

openalex-database