The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user or agent to download the affine-mcp-server package from the npm registry and pull Docker images from GitHub Container Registry (ghcr.io/dawncr0w/affine-mcp-server). These are third-party resources not maintained by a recognized trusted vendor.
[REMOTE_CODE_EXECUTION]: The installation instructions utilize npm i -g and npx, which execute code from the remote npm registry on the local system. This is a standard vector for supply chain attacks if the package were to be compromised.
[COMMAND_EXECUTION]: The skill provides instructions for running CLI commands (affine-mcp, docker run) and configuring AI clients to execute the affine-mcp binary to start the server. This grants the agent the ability to spawn local processes.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to read and manipulate external data (AFFiNE documents, databases, and comments). This creates a vulnerability surface where malicious instructions embedded in a document could be interpreted by the AI agent.
Ingestion points: Data is pulled into the agent's context through tools like read_doc, search_docs, list_comments, and inspect_template (SKILL.md).
Boundary markers: The skill documentation does not mention the use of delimiters or explicit warnings to the agent regarding untrusted content in documents.
Capability inventory: The skill provides significant capabilities to the agent, including the ability to create, update, and delete documents and databases (create_doc, update_doc, delete_doc, delete_workspace).
Sanitization: No specific sanitization or filtering logic is described for content retrieved from the AFFiNE API.

affine-mcp-server-integration