The Agent Skills Directory

[SAFE]: The skill is focused on DFM analysis with no evidence of malicious patterns, obfuscation, or unauthorized data access. The instructions and scripts are well-documented and align strictly with the stated purpose of assisting engineers in manufacturing reviews.
[EXTERNAL_DOWNLOADS]: The skill uses standard Python packages (trimesh, numpy, rtree, scipy, networkx, lxml) listed in requirements.txt. These packages are sourced from well-known registries and are standard industry tools for geometric calculations.
[COMMAND_EXECUTION]: The script scripts/mold_tool.py is executed locally to process 3D mesh data. It performs purely mathematical analysis of geometry facts without performing network operations, accessing sensitive system files, or attempting privilege escalation.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user 3D models for analysis. Ingestion points: Processes .stl, .obj, .ply, and .3mf files via scripts/mold_tool.py. Boundary markers: Uses a structured Markdown report format for outputs as defined in SKILL.md. Capability inventory: Limited to geometric calculation and data reporting; no capabilities for file modification or network communication are present in the measurement tools. Sanitization: Utilizes the trimesh library to parse geometric primitives and drop degenerate faces, effectively isolating the processed data from the LLM's natural language instruction context.

dfm