The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts to perform code health audits and score research papers. Specifically, it runs scripts/wiring_sweep.py to analyze the codebase and references dynamic_scorer.py from the associated research-scanner skill. These operations are restricted to the local environment and intended for development auditing.
[EXTERNAL_DOWNLOADS]: The pipeline leverages the scholarx MCP tool to perform search and bulk download operations for research papers (e.g., from arXiv). This data is ingested into a local Knowledge Graph for comparative analysis, which is the primary documented purpose of the skill.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes external research papers to generate development plans.
Ingestion points: External papers are downloaded via mcp_scholarx_sx_storage and ingested using mcp_agent-utilities-kg_kg_ingest.
Boundary markers: None explicitly defined in the orchestration steps to delimit untrusted paper content from system instructions.
Capability inventory: The skill has capabilities to write to the Knowledge Graph and execute local Python scripts for analysis.
Sanitization: No specific content sanitization or filtering is described for the paper extraction process. Given the research-focused use case, this is categorized as a low-risk architectural characteristic rather than a malicious finding.

agent-utilities-evolution