The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data that could theoretically contain instructions designed to influence the agent's behavior.
Ingestion points: The skill instructs the agent to fetch documentation from https://mastra.ai/llms.txt and various subpages on https://mastra.ai (as described in references/remote-docs.md).
Boundary markers: There are no specific delimiters or warnings instructed to separate the fetched documentation from the agent's primary instruction set.
Capability inventory: The agent is granted extensive capabilities including file system operations (writing TypeScript files, creating project directories, reading node_modules), shell command execution (npm install, npx, grep, docker), and network access (web fetching).
Sanitization: No sanitization or validation protocols are defined for the content retrieved from the remote documentation URLs.
[COMMAND_EXECUTION]: The skill's workflow depends on executing numerous shell commands to manage a development environment.
Evidence includes instructions to use npm, npx, grep, ls, and cat for project setup, dependency management, and documentation lookup.
[EXTERNAL_DOWNLOADS]: The skill frequently references and retrieves content from external sources.
It fetches documentation files from the official framework domain (https://mastra.ai) and instructs the installation of packages from the official NPM registry.

mastra