The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process documentation from external sources, which represents an indirect prompt injection surface. The ingestion points include local documentation files within node_modules/@mastra/*/dist/docs/ and remote content fetched from the official mastra.ai website. While instructions direct the agent to verify memory against these docs, there is no explicit sanitization for the retrieved content. The capability inventory includes ls, cat, grep, and curl for retrieval, plus the mastra api CLI for resource interaction.\n- [COMMAND_EXECUTION]: The skill documents numerous shell commands for project setup (npm create mastra), dependency management, and system operations (ls, docker run). It also provides detailed instructions for the mastra api CLI, which executes commands to interact with agent and workflow resources on local and remote servers.\n- [EXTERNAL_DOWNLOADS]: The skill fetches documentation and interacts with platform APIs at mastra.ai, observability.mastra.ai, and output.signals.mastra.ai. These are recognized as legitimate vendor resources for the Mastra framework.

mastra