mantis-structural-index
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFE
Full Analysis
- Dynamic Script Generation: The skill generates and executes Python helper scripts (
build_structural_index.pyandquery_structural_index.py) at runtime. These scripts manage the indexing and querying processes, using a version-marker system to ensure consistency and allow for regeneration when updates are needed. - File System Access Controls: It adheres to a strictly defined input/output contract, reading from the source root and writing index data to specific workspace paths. It implements 'Sentinel Checks' to ensure the code root matches the expected snapshot and prevents write operations on source files when they are in a pinned state.
- Sub-Agent Communication Pattern: The instructions specify that the skill should not interact with the user directly, as it is designed to function as a backend component for orchestration agents. This operational silence is a design characteristic of the pipeline architecture.
- Semantic Tool Integration: The skill integrates with well-established semantic analysis backends (e.g., LSIF, SCIP, Kythe) and fallback utilities (e.g., ctags, tree-sitter). This allows it to scale its precision based on the tools available in the host environment.
Audit Metadata