The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses curl to fetch the content of external webpages to automatically generate titles for reference links. This constitutes an indirect prompt injection surface where untrusted data from the internet enters the agent's context.
Ingestion points: External URLs provided during the blog writing process.
Boundary markers: The skill does not specify delimiters or instructions to ignore embedded commands within the fetched webpage titles.
Capability inventory: The agent has extensive capabilities including file system modification, command execution (git, gh, zola), and network access.
Sanitization: No explicit sanitization or validation of the remote content is performed before it is integrated into the post.
[COMMAND_EXECUTION]: The skill executes various system commands and CLI tools, including git for version control, gh for managing pull requests, and zola for building the static site. It also executes a local repository script ./switch-site.sh to configure the environment.
[EXTERNAL_DOWNLOADS]: The skill clones source code and content from GitHub repositories (jim60105/blog and bot0419/ai-talks-content). It also provides instructions for downloading and installing the Zola binary from its official GitHub releases page.

create-blog-post