The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires installation of the @jackwener/opencli NPM package and the 'Playwright MCP Bridge' Chrome extension, which originate from third-party sources outside of established trusted organizations.
[COMMAND_EXECUTION]: The skill operates by executing shell commands via the opencli binary and the playwright MCP server to interact with target websites.
[REMOTE_CODE_EXECUTION]: The 'Self-iteration' capability instructs the agent to create new adapters by writing YAML files to ~/.opencli/clis/. These files contain arbitrary JavaScript code within evaluate blocks that the tool subsequently executes, enabling dynamic code execution.
[DATA_EXFILTRATION]: The skill is designed to access sensitive personal data across multiple platforms, including private history, bookmarks, and potentially messages (via Playwright fallback), by leveraging the user's authenticated browser sessions.
[PROMPT_INJECTION]: The skill processes untrusted content from social media platforms and search results, presenting a surface for indirect prompt injection attacks.
Ingestion points: Data retrieved from external sites via opencli and Playwright snapshots.
Boundary markers: None are specified in the instructions to protect against instructions embedded in the ingested data.
Capability inventory: Shell command execution, local file modification, and browser automation.
Sanitization: No sanitization of ingested content is defined before it is processed by the agent.

opencli