The Agent Skills Directory

[COMMAND_EXECUTION]: The utility script catalog-endpoints.sh uses standard tools for file discovery and content analysis. It features defensive checks for input existence and readability, resolves absolute paths, and prevents path traversal through pattern matching.
[INDIRECT_PROMPT_INJECTION]: The skill processes existing source code to generate new resolvers and schema definitions, creating a surface for indirect instructions to influence the agent.
Ingestion points: The agent reads project source files identified during the cataloging phase.
Boundary markers: The instructions do not define specific delimiters to isolate the ingested source code from the agent's internal reasoning context.
Capability inventory: The agent can write new source files and execute package manager commands.
Sanitization: The skill's security rules require validation of JSON and SDL formatting prior to saving generated content.
[CREDENTIALS_UNSAFE]: The skill contains strict rules against the exposure of secrets and provides structured guidance to ensure that existing API security controls are correctly ported to the new GraphQL interface.
[EXTERNAL_DOWNLOADS]: The documentation references installation from the author's repository using npx.

rest-to-graphql-deriver