The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/cookies.ts launches the Google Chrome browser using child_process.spawn. It uses the Chrome DevTools Protocol (CDP) to automate the browser and extract authentication cookies. While this is scoped to the skill's specific data directory, it represents an automated interaction with local software.
[DATA_EXPOSURE]: The skill manages sensitive authentication data by reading and writing to local files. It stores session cookies and user consent records in the user's application data directory (e.g., ~/.local/share/baoyu-skills/ or ~/Library/Application Support/baoyu-skills/).
[INDIRECT_PROMPT_INJECTION]: The skill serves as a data ingestion point, fetching content from X (tweets and articles) and converting it into Markdown. Because this content originates from external users, it could contain malicious instructions designed to influence the behavior of the AI agent when it processes the resulting Markdown files.
Ingestion points: Fetches data from x.com via GraphQL and downloads media from twimg.com (scripts/graphql.ts, scripts/media-localizer.ts).
Boundary markers: The output uses YAML front matter to separate metadata, but the body consists of raw external text.
Capability inventory: The skill has capabilities to write files, read local data files, and perform network operations.
Sanitization: Basic Markdown escaping is performed in scripts/markdown.ts, but there is no specific filtering for prompt injection patterns in the ingested text.

baoyu-danger-x-to-markdown