Skip to main content
Glama

WeaveGraph

Semantic Code-Graph Worktree Scheduler, Min-Cut Multi-Agent Task Partitioner & 3-Way AST Semantic Merge Verifier

WeaveGraph solves the #1 bottleneck identified across the Global GitHub AI-500 & Complexity-100 Atlas (AI Coding Agents & Dev Harnesses, 121 repos): parallel coding agents (Claude Code, Codex, OpenCode, Orca) running in git worktree lanes frequently edit coupled files and pass text-based git merge cleanly while silently breaking function signatures, keyword parameters, imports, or schema fields.

Key Capabilities

  1. Native C++20 CSR Graph Engine (csrc/weavegraph_kernels.cpp):

    • Iterative Tarjan Strongly Connected Components ($O(V+E)$)

    • Shortest-hop Reverse-Dependency Blast Radius BFS

    • Fiduccia-Mattheyses / Kernighan-Lin Min-Cut Partitioning minimizing cross-agent coupling across $K$ parallel git worktrees

    • Topological Dependency Wave Scheduling over condensed SCC DAGs

  2. 3-Way Semantic AST Merge Verifier (verify-merge):

    • Catches 5 classes of silent cross-agent contract breaks invisible to git merge: SIGNATURE_ARITY_BREAK, REMOVED_KEYWORD_PARAM, DANGLING_IMPORT_OR_SYMBOL, CONTRACT_FIELD_REMOVED, and DUPLICATE_DEFINITION_CLASH.

  3. Drop-In CLI, FastAPI REST Server & Stdio MCP Server (.mcp.json):

    • Exposes weavegraph_index_repo, weavegraph_blast_radius, weavegraph_plan_worktrees, and weavegraph_verify_merge to Claude Code, Cursor, and Codex.

  4. Interactive React 18 + TypeScript Workbench (ui/):

    • Visualizes worktree lanes, cross-agent coupling reduction, semantic merge violations, and blast-radius hops.

Related MCP server: Cordyceps Search

Quickstart

uv sync --python 3.13 --all-extras
uv run weavegraph demo
uv run weavegraph scan .
uv run weavegraph plan . --agents 3
uv run weavegraph serve --port 8150

Verification & Quality Gates

uv run ruff format --check src tests
uv run ruff check src tests
uv run mypy src tests
uv run pytest --cov=weavegraph --cov-report=term-missing --cov-fail-under=95
cd ui && npm ci && npm run build

Author, LinkedIn & Enterprise AI Architecture Hub

Built by Nagarjuna Reddy Kamireddy (Nagarjuna Kamireddy / Nagarjuna K) — Lead AI Architect | Production Agentic AI, Hybrid RAG & LLM Evals (750K+ Workflows/Yr) | Ex-John Deere, GSK, Deloitte | AWS SA Pro & NVIDIA Agentic AI Certified.

Available Tools

4 tools
weavegraph_blast_radiusC

Compute shortest-hop transitive blast radius from changed seed files.

Inputs: repo_path (required), seed_files (required).

weavegraph_index_repoC

Index a repository AST dependency graph, Tarjan SCC cycles, and topological build waves.

Inputs: repo_path (required), max_files.

weavegraph_plan_worktreesC

Partition files and tasks across K parallel coding agent git worktrees using min-cut graph optimization.

Inputs: repo_path (required), tasks, num_agents.

weavegraph_verify_mergeC

Verify 3-way merge safety across Base, Agent A, and Agent B file maps for silent semantic AST breaks.

Inputs: base_tree (required), agent_a_tree (required), agent_b_tree (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.3/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool targets a clearly distinct operation: indexing, merge verification, blast-radius computation, and worktree planning. There is no meaningful overlap in purpose or likely inputs/outputs.

    Naming Consistency4/5

    All tools use the same weavegraph_ snake_case prefix and mostly follow a verb_noun pattern. The exception is weavegraph_blast_radius, which omits an explicit verb such as compute_.

    Tool Count5/5

    Four tools is a tight, well-scoped surface for this specialized graph-analysis workflow. Each tool earns its place without obvious redundancy.

    Completeness4/5

    The set covers indexing, impact analysis, merge safety, and parallel worktree planning, which is strong lifecycle coverage. It lacks a direct tool for querying or inspecting the indexed dependency graph beyond derived analyses.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables agentic coding tools to maintain a living, in-repo model of intent, architecture, and code, with deterministic indexing, bidirectional linking, and continuous drift detection for task-scoped context and reconciliation.
      MIT
    • F
      license
      Not graded
      quality
      B
      maintenance
      Enables developer agents to perform semantic codebase search, dependency and impact analysis, cross-file refactoring, and full-stack API tracing through a unified query DSL over a high-performance graph engine.
      -
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI coding agents to query a local semantic code graph for surgical context, cross-file resolution, framework-aware routes, and code intelligence while keeping indexes automatically up to date.
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Enables AI coding agents to index a workspace into AST symbol graphs and retrieve exact function or class slices, trace callers and blast radius, and apply syntax-safe symbol patches without dumping whole files into context.
      7
      1
      MIT