The Agent Skills Directory

[COMMAND_EXECUTION]: The skill guides the agent to use the ontoly CLI for building and analyzing software graphs. Commands such as ontoly build, ontoly coverage, and ontoly stats are used to process the local repository.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the target repository, including source code files and the generated .ontoly/SoftwareGraph.json graph. This represents an attack surface for instructions embedded in the data being analyzed.
Ingestion points: The agent reads the local .ontoly/SoftwareGraph.json file and inspects source files via InspectFile or InspectModule tools.
Boundary markers: While the skill mandates citing evidence and reporting confidence levels, it does not specify explicit delimiters or sanitization steps for the ingested code content.
Capability inventory: The skill utilizes the ontoly CLI and various MCP lookup tools (FindNode, InspectFile, etc.) to analyze the repository.
Sanitization: No specific filtering or escaping mechanisms for the analyzed file content are described.

dependency-analysis