The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill relies on bunx cosense, which downloads and executes the cosense package from the npm registry at runtime.
[COMMAND_EXECUTION]: The skill operates by executing shell commands via the bunx runner to interact with the Cosense service for fetching, searching, and creating wiki pages.
[CREDENTIALS_UNSAFE]: Authentication is performed using a session ID (connect.sid). The skill documentation instructs users to set this token via command-line arguments (e.g., bunx cosense profile set personal --sid <connect.sid>), which can expose the secret in shell history and process monitoring tools.
[PROMPT_INJECTION]: The skill retrieves arbitrary text content from external wiki pages, creating a surface for indirect prompt injection where malicious instructions could influence the agent's behavior.
Ingestion points: The page get, page search, and export commands fetch content from the external Cosense service into the agent's context.
Boundary markers: No explicit delimiters or instructions are used to separate external wiki content from the agent's internal logic.
Capability inventory: The skill possesses the ability to execute shell commands and modify wiki content based on the data it processes.
Sanitization: No sanitization or validation mechanisms are described for the content retrieved from external sources.

cosense