The Agent Skills Directory

[DATA_EXFILTRATION]: The skill manages credentials in standard locations such as ~/.heygen/credentials and ~/.aws/credentials for authenticating cloud rendering and infrastructure deployment. The feedback command transmits diagnostic reports to an external channel; the skill documentation mandates the redaction of absolute paths, machine identifiers, and secrets before submission.
[EXTERNAL_DOWNLOADS]: Fetches pinned Chromium binaries, machine learning models for transcription (Whisper) and text-to-speech, and configuration files from the vendor's official GitHub repositories and domains.
[INDIRECT_PROMPT_INJECTION]: The capture command ingests content from external URLs to generate project scaffolds (Ingestion: references/init-and-scaffold.md). The tool generates a BLOCKED.md file to identify restricted content (Boundaries: present). The skill maintains high capabilities including file writing and cloud deployment (Capabilities: render, publish, lambda deploy). Documentation provides warnings against using data from failed captures (Sanitization: present).
[COMMAND_EXECUTION]: Orchestrates various development and deployment tasks by executing shell commands via npx and integrating with system tools like docker, terraform, gcloud, and aws-sam.

hyperframes-cli