The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use curl within terminal environments to upload media bytes to pre-signed URLs provided by internal tools.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could potentially contain malicious instructions intended to influence the agent. Ingestion points: External media via media_from_url, user-provided files via media_from_file, and style guidelines from ~/.uplika/<workspace-slug>/voice.md. Boundary markers: Step 6 requires the agent to show the exact text and media to the user and obtain manual approval before publishing, serving as a critical safety boundary. Capability inventory: The agent can perform network requests to publishing APIs, execute shell commands (curl), and delete existing posts. Sanitization: The procedure relies on manual human verification of the final post content rather than automated programmatic sanitization.
[EXTERNAL_DOWNLOADS]: The skill facilitates the retrieval of media from external HTTPS URLs provided by the user for attachment to social media posts.

publish