The Agent Skills Directory

[COMMAND_EXECUTION]: The skill guides the agent to use npx tsc and npx convex dev for code verification and deployment. These commands are standard for the framework and are used for their intended purpose.- [EXTERNAL_DOWNLOADS]: The skill recommends the use of official vendor packages such as @convex-dev/agent and @convex-dev/workflow for implementing specific features.- [INDIRECT_PROMPT_INJECTION]: The agent is instructed to read project-specific files like convex/schema.ts and generated guidelines. While this represents a potential ingestion point for untrusted data, it is a necessary part of the backend development workflow.
Ingestion points: Reads convex/schema.ts and convex/_generated/ai/guidelines.md (SKILL.md).
Boundary markers: None specified.
Capability inventory: Executes shell commands (npx) and writes files to the convex/ directory (SKILL.md).
Sanitization: No explicit sanitization of input files is mentioned.

convex-expert