The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches documentation, API conventions, and configuration guidelines from established and well-known services, including Saleor's official documentation (docs.saleor.io), The Guild (the-guild.dev), and GraphQL's official site (graphql.org).
[INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface where external web content is ingested to inform code generation.
Ingestion points: The agent is instructed to use WebSearch and WebFetch to retrieve live documentation from Saleor and The Guild in SKILL.md.
Boundary markers: None provided in the skill instructions.
Capability inventory: The skill has access to Write, Edit, and Bash tools, which could potentially be used to create or execute files based on fetched content.
Sanitization: No specific sanitization or validation of the fetched documentation is mentioned.
[SAFE]: The skill's instructions focus on standard development practices such as using variables, fragment reuse, and type generation. It encourages secret management best practices by recommending the use of environment variables or headers for authorization tokens rather than hardcoding them.

graphql-dev