The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses system commands to handle environment-specific tasks and browser automation.
scripts/paths.ts utilizes execSync to run cmd.exe /C "echo %USERPROFILE%" for resolving Windows home paths when running inside WSL.
scripts/cookies.ts uses spawn to launch a Chrome or Chromium browser instance to facilitate automated cookie extraction through the Chrome DevTools Protocol (CDP).
[EXTERNAL_DOWNLOADS]: The skill fetches data and assets from official X (Twitter) domains to perform its primary function.
scripts/media-localizer.ts downloads image and video assets from pbs.twimg.com and video.twimg.com to local storage.
scripts/graphql.ts fetches API responses and JavaScript bundles from x.com and abs.twimg.com to extract GraphQL query identifiers.
[PROMPT_INJECTION]: The skill processes external tweet content, establishing an indirect prompt injection surface.
Ingestion points: Tweet text and article content are ingested from the X API and formatted into markdown files.
Boundary markers: The skill uses YAML front matter and Markdown headers to structure the output, though it does not provide explicit instructions to the agent to ignore instructions embedded in the tweet body.
Capability inventory: The skill possesses capabilities for file system writes, network requests to X, and spawning browser processes.
Sanitization: It implements slug sanitization for file paths and escapes specific characters in media metadata to prevent structural markdown injection.
[CREDENTIALS_UNSAFE]: The skill contains a hardcoded bearer token for API access.
scripts/constants.ts defines a DEFAULT_BEARER_TOKEN. This is a publicly known guest bearer token used by the X web client and is required for guest-level API interactions.

tuzi-danger-x-to-markdown