The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local shell execution for curl, node, and python3 to manage API requests, execute helper scripts, and handle text encoding conversions. These tools are used to interact with the IMA backend and ensure data integrity across different operating systems.
[EXTERNAL_DOWNLOADS]: The skill directs network traffic to ima.qq.com and Tencent Cloud's myqcloud.com. The use of myqcloud.com is required for uploading files to the IMA knowledge base via Tencent Cloud Object Storage (COS), although this domain is not explicitly listed in the allowed_domains metadata field.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its document-processing capabilities. Ingestion points: The skill reads user notes via get_doc_content and knowledge base content via search_knowledge and get_knowledge_list. Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat ingested data as untrusted. Capability inventory: The skill can perform shell command execution, network operations, and local file access. Sanitization: No escaping, filtering, or validation is performed on the content retrieved from external sources before it is processed by the agent.

ima-skills