go-code

command
v1.39.1 Latest Latest
Warning

This package is not in the latest version of its module.

Go to latest
Published: Jul 17, 2026 License: Apache-2.0 Imports: 98 Imported by: 0

Documentation

Overview

cmd/go-code/http_resolve.go

go-code — Code intelligence MCP server.

Provides multi-language AST parsing via tree-sitter, repository analysis, code comparison, and dependency graph visualization. Runs as HTTP MCP server (default) or stdio transport (--stdio flag).

Tools: repo_analyze, file_parse, code_compare, dep_graph, symbol_search, call_trace, impact_analysis, dead_code, explore, code_health, code_graph, repo_search, code_search, semantic_search

cmd/go-code/tool_debug_investigate.go

cmd/go-code/tool_debug_investigate_alerts.go

cmd/go-code/tool_debug_investigate_body.go

cmd/go-code/tool_debug_investigate_deadcode.go

cmd/go-code/tool_debug_investigate_downstream.go

Sprint B4: downstream callgraph walk for root-cause discovery.

Motivation: B2 (upstream) walks callers UP from the symptom site. But for cases like SFU session.rs:226 (event-loop run) → bug is in accept_renegotiation_answer (renegotiation.rs:363) which the loop CALLS, we need to walk DOWN. B4 is symmetric to B2 but follows callees.

cmd/go-code/tool_debug_investigate_fleet.go

cmd/go-code/tool_debug_investigate_format.go

cmd/go-code/tool_debug_investigate_fusion.go

cmd/go-code/tool_debug_investigate_history.go Phase γ.C — historical incidents persistence + retrieval.

cmd/go-code/tool_debug_investigate_impact.go

cmd/go-code/tool_debug_investigate_llm.go

cmd/go-code/tool_debug_investigate_metrics.go

cmd/go-code/tool_debug_investigate_symbolbody.go

cmd/go-code/tool_debug_investigate_symbols.go

cmd/go-code/tool_debug_investigate_traces.go

cmd/go-code/tool_debug_investigate_upstream.go

Sprint B2: upstream callgraph walk for root-cause discovery.

Motivation (dogfooding gap): Phase 3 attributed the anomaly to session.rs:226 (the UDP send site) but the actual root cause was accept_renegotiation_answer (the caller's caller). B1 gives the LLM the function body; B2 ensures the upstream callers are in the hypothesis pool BEFORE FusionRank, so the rank pipeline (recency + complexity + impact) can promote the actual culprit.

cmd/go-code/tool_resolve_frame.go

Source Files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL