The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes several local Python scripts to perform its core functions. These include 'scripts/dfs_client.py' for JS-rendered analysis, 'scripts/google_client.py' for status inspection, 'scripts/schema_gen.py' for code generation, and 'scripts/report_publish.py' for generating final deliverables.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface. * Ingestion points: The skill fetches HTML, JSON-LD, Microdata, and RDFa content from user-specified external URLs via the 'webfetch' tool. * Boundary markers: There are no specific instructions or delimiters mentioned to isolate the external web content or prevent the agent from following instructions embedded within the fetched data. * Capability inventory: The agent has access to multiple scripts ('dfs_client.py', 'google_client.py', 'schema_gen.py', 'report_publish.py') which execute shell commands to process data and publish reports. * Sanitization: The skill description does not specify any sanitization, filtering, or validation of the fetched external HTML content before it is processed by the agent or passed to the underlying scripts.

schema-validator