Athena: A Local-First Git & AST Change-Intelligence Engine and MCP Server A developer released Athena (athena-mcp), an open-source, local-first Git and AST change-intelligence engine, CLI, and Model Context Protocol server that parses Python and TypeScript/JavaScript structural contracts across Git revision ranges to detect breaking changes, trace downstream caller blast radius, and flag stale documentation. Each finding is bound to a verified EvidenceRecord containing the 40-character commit SHA, file path, base/target line spans, and diff hunk excerpt, and the tool runs entirely on open-weights models (Gemma 7B and Qwen 2.5 Coder 7B via Ollama) on consumer hardware, tested on a 6 GB VRAM RTX 4050 laptop GPU. A verifier strips fabricated commit SHAs or file paths and falls back to deterministic analysis, and the project ships stdio MCP tools including inspect_revision_range, analyze_code_changes, detect_potential_breaking_changes, and generate_release_notes. This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle https://dev.to/challenges/hacktoberfest-week2-2026-10-12 Most release-note generators and PR review bots do the same thing: they dump git log --oneline into a cloud LLM and hope for the best. That fails in real engineering teams for two reasons: fix api : minor cleanup in parameter defaults might actually insert an optional parameter before an existing one def compute a, b=2 - def compute a, inserted=0, b=2 , silently shifting positional arguments for every caller in production. Athena athena-mcp is an open-source, local-first Git & AST Change-Intelligence Engine, CLI, and Model Context Protocol MCP server . Instead of guessing from commit titles, Athena parses the actual Python Abstract Syntax Tree ast and TypeScript/JavaScript structural contracts across any Git revision range base ref..target ref , traces downstream caller blast radius across the repository, flags stale Markdown documentation, and binds every single finding to a verified EvidenceRecord 40-character commit SHA, file path, exact base/target line spans, and diff hunk excerpt . Because this challenge celebrates Open-Source AI , Athena is built 100% on open-weights models and open protocols that run locally on consumer hardware tested on an NVIDIA RTX 4050 Laptop GPU with 6 GB VRAM and 16 GB RAM : gemma:latest Gemma 7B — Open Weights via Ollama, ~5.0 GB : qwen2.5-coder:7b Qwen 2.5 Coder 7B — Open Weights via Ollama, ~4.7 GB : src/athena/ollama engine.py : /api/ps endpoint before inference and automatically evicts any competing loaded model keep alive: 0 before loading src/athena/verifier.py : gemma:7b or EvidenceRecord table. If the model invents a commit SHA, fabricates a file path, or times out, Athena strips the unverified claim or gracefully falls back to deterministic mcp Python SDK : stdio MCP tools inspect revision range , analyze code changes , detect potential breaking changes , and generate release notes so any MCP-compatible agent or IDE can query ground-truth Git + AST diffs locally. None -fallback body guards griffe 469 , str, Enum <- StrEnum breaking migrations griffe 501 , and newly raised exceptions griffe 498 . BUG-030 : old pos idx n = new pos idx n , verified against a 66-case parametrized binding-relation matrix across positional-only / , standard positional, trailing appends, and keyword-only controls. BUG-021 .. BUG-029 : NameError s. all , init .py re-exports, and TypeScript export symbols is public api export=True from internal private modules. getattr , globals , eval , importlib.import module , TS bracket dispatch obj key and outputs an explicit AnalysisCoverageReport . .github/workflows/athena-audit.yml or your IDE Claude Desktop , Cursor , VS Code via MCP. Python 3.12+ bash git clone https://github.com/aaryanrwt/athena.git cd athena pip install -e ". dev "