The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it processes untrusted external data (uploaded files and search results) which is then used to generate answers or guide further retrieval steps.
Ingestion points: File uploads via mxbai.stores.files.upload (e.g., guide.pdf, faq.md in SKILL.md) and search results retrieved through mxbai.stores.search and mxbai.stores.question_answering.
Boundary markers: The provided instructions do not specify the use of delimiters or explicit warnings to the agent to ignore instructions embedded within the processed external data.
Capability inventory: The skill uses the Mixedbread SDK to perform network operations and multi-step retrieval through its API.
Sanitization: No sanitization or validation of the content within the uploaded documents or retrieved search chunks is described.
[EXTERNAL_DOWNLOADS]: The skill documentation references the installation of external software packages required for operation.
Python: mixedbread package.
Node.js: @mixedbread/sdk package.
These are the official SDKs associated with the skill author.

mixedbread-search