The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to manage and process content from Markdown (.md), MDX (.mdx), and JSON files. This creates an attack surface where instructions embedded in the content could potentially influence the agent's behavior during editing or retrieval tasks.\n
Ingestion points: The agent is instructed to read content files from repository paths such as content/posts/, content/docs/, and content/pages/ as defined in the provided tina/config.ts templates.\n
Boundary markers: The skill templates do not provide explicit boundary markers or instructions for the agent to ignore embedded commands within the content files.\n
Capability inventory: The skill leverages Bash for environment setup and Write/Edit for configuration management, which could be targeted by an injection.\n
Sanitization: There is no evidence of content sanitization to prevent the interpretation of embedded instructions.\n- [COMMAND_EXECUTION]: The skill documentation and initialization scripts (scripts/check-versions.sh) involve executing shell commands to manage packages and check versions. This includes global installations (npm install -g pnpm) and executing the TinaCMS CLI via npx.\n- [EXTERNAL_DOWNLOADS]: The skill facilitates the download of tools and dependencies from the official NPM registry and GitHub repositories. These downloads are directed toward well-known, trusted organizations associated with TinaCMS, Cloudflare, and React.

tinacms