Neural network background
Source: packages/fabric-mcp/src/tools.tsmcpgenerated

fabric_query

Ask "what symbols match this name and what do they connect to?" Matches nodes by name (case-insensitive substring + optional Damerau–Levenshtein fuzzy), then walks edges up to depth hops. Returns typed nodes (FN/TYP/IFC/MOD/TST/CONTRACT/DEBT/GOV/DEP) plus the edges touching them. Use over grep when you need relationship structure, not just text matches.

Input schema

FieldTypeRequiredDescription
querystringyesSubstring to match against node names. Case-insensitive.
depthnumbernoEdge traversal depth from each matched seed. 1 = immediate neighbours. (default: 1)
fuzzybooleannoAllow edit-distance matches for typos (e.g. "authentcate" finds "authenticate"). Ranked by score; exact substring hits always win ties. (default: false)
fuzzy_thresholdnumbernoMinimum fuzzy score (0.0–1.0) for a node to be included when fuzzy=true. Default 0.6.
max_resultsnumbernoCap on returned nodes. (default: 20)
typesarraynoRestrict to one or more node types: FN, TYP, IFC, MOD, TST, DEP, DEBT, CONTRACT, GOV. (items: string)

Source

Source: packages/fabric-mcp/src/tools.ts