The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses files.write to generate Python scripts and python.run to execute them for generating PNG and DXF patent drawings. This runtime script generation and execution represents a dynamic execution surface.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted technical data from user input (PATENT_INPUT) and external web search results (mcp_web-search__web_search). This data influences the generation of reports and executable Python code.
Ingestion points: PATENT_INPUT variable, results from mcp_web-search__web_search, and text from mcp_patent-search__patsnap_fetch.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the processing logic for external data.
Capability inventory: The skill utilizes python.run, files.write, and various MCP tools to interact with the filesystem and external services.
Sanitization: There is no evidence of sanitization or validation of the ingested patent text or web content before it is processed or used in code generation.
[EXTERNAL_DOWNLOADS]: The skill invokes runtime.apply_sync to ensure the installation of external Python libraries including ezdxf, matplotlib, and numpy. While these are well-known libraries, the automated installation of dependencies at runtime is a notable behavior.
[COMMAND_EXECUTION]: The skill relies on executing system-level operations via the python.run and files.write tools to fulfill its primary purpose of report and drawing generation.

patent-lifecycle-agent