The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses an inline Node.js script (node -e) to dynamically process MDX content and generate JSON payloads at runtime. This script performs file system operations (reading source files and writing to /tmp) and string manipulation to prepare data for the database.
[INDIRECT_PROMPT_INJECTION]: The skill ingest and processes external MDX files, creating a potential surface for instructions hidden in data to influence agent behavior.
Ingestion points: MDX content is read from local file paths (e.g., ~/Code/joelhooks/joelclaw/apps/web) provided by the user or identified by the agent.
Boundary markers: No explicit delimiters or instructions are used to signal that the agent should ignore instructions embedded within the processed content.
Capability inventory: The skill possesses shell execution capabilities (npx convex) and network access (curl), which are accessible during the content processing lifecycle.
Sanitization: Processing is limited to frontmatter removal and does not include filtering for potentially malicious natural language instructions.
[COMMAND_EXECUTION]: The skill executes shell commands using npx convex for database interactions and curl for triggering site revalidation and verifying deployment status.
[EXTERNAL_DOWNLOADS]: The skill uses the npx command, which may download and execute the convex CLI package from the public registry if it is not locally available.

content-publish