The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the @cargo-ai/cli package from the npm registry and allows adding further functionality via npx skills add. These resources originate from the skill's authoring vendor.- [COMMAND_EXECUTION]: The skill utilizes several CLI tools including npm, grep, gh, and the vendor's own cargo-ai tool for environment setup, data retrieval, and repository interaction.- [DATA_EXPOSURE_AND_EXFILTRATION]: The instructions include reading the agent's local configuration file ~/.claude/plugins/installed_plugins.json to check for existing installations and sending session metadata to the vendor's workspace management API for attribution.- [PERSISTENCE_MECHANISMS]: The skill creates a persistent marker file at ~/.config/cargo-ai/.star-asked to track whether the user has been prompted to star the repository, ensuring the request is not repeated in future sessions.- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources such as LinkedIn and company websites which may contain untrusted content.
Ingestion points: Data enters through the enrichCompanyFromDomain, extract, and searchJobs actions via the cargo-ai CLI.
Boundary markers: Explicit instructions command the agent to categorize every line as 'Sourced', 'Inferred', or 'Unknown' and strictly prevent blending different data types.
Capability inventory: The agent has access to network operations via the vendor CLI and gh tool, and can write to the local file system.
Sanitization: The skill requires the agent to report data gaps clearly as 'Unknown' and prohibits the fabrication of priorities or competitor information.

research-account