Wisegit

wisegit

[npm] [MCP Registry] [License: MIT]

"Don't take a fence down until you know the reason it was put up." — G.K. Chesterton

wisegit is a local MCP server that extracts decision intent from git history and protects intentional code from AI modification.

When Claude Code (or any MCP-compatible agent) is about to edit a file, wisegit injects a decision manifest showing which functions are frozen, stable, or open — so the AI respects what was intentional, not just what compiles.

Zero config. Zero external services. Everything local.

Install

# Set up any repo (one command)npx @sandip124/wisegit setup
# Or add as MCP server globallyclaude mcp add wisegit -- npx @sandip124/wisegit serve

Published on

• npm: @sandip124/wisegit
• MCP Registry: io.github.Sandip124/wisegit
• GitHub: Sandip124/wisegit

The Problem

LLMs have no concept of intentional code. A manually-tested fix and a broken stub look identical — both are just text. Real scenario:

1. You fix a Stripe race condition with sleep(350) — manually tested, committed.
2. Next session: "find bugs." Claude removes sleep(350) — looks like dead code.
3. Production incident.

Root cause: git history contains proof of intention. Nobody extracts it.

How It Works

Git History → Tree-sitter AST → Intent Extraction → SQLite Event Store → MCP Tools

1. Indexes your git history — walks every commit, parses diffs at the AST level (function boundaries, not line counts)
2. Classifies commits — STRUCTURED (fix:, feat:), DESCRIPTIVE (plain sentences), or NOISE (wip, x)
3. Extracts intent — rule-based for structured/descriptive commits, LLM for noise (Phase 2)
4. Computes freeze scores — 0–1 per function, derived from protection signals (git, issue, code structure, test, structural, Naur, Aranda) minus adaptive obsolescence penalty (8 signals, entropy-calibrated). Age uses Weibull survival [18]; expertise uses DOE model [19]
5. Serves decision manifests via MCP — Claude Code calls get_file_decisions before editing any file

What the AI Sees

[DECISION MANIFEST: payment.service.cs]━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━FROZEN:  ProcessPayment()  [score: 0.89] [Recovery: L1]  - sleep(350) → Stripe race condition. Won't Fix.    HIGH — commit a3f19b2
STABLE:  ValidateOrder()  [score: 0.55] [Recovery: L2]  - Fixed null reference on Safari iOS WebKit.    MEDIUM — commit 7c14694
OPEN:    FormatReceipt()  [score: 0.12] [Recovery: L3]  ← safe to modify━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

• FROZEN (score >= 0.80): Do not modify without explicit user approval
• STABLE (score 0.50-0.79): Proceed with caution, review intent first
• OPEN (score < 0.50): Safe to modify freely

Quick Start

Prerequisites

• Node.js >= 20

That's it. No Docker, no PostgreSQL, no external services.

1. Set Up a Repository (one command)

cd /path/to/your/reponpx @sandip124/wisegit setup

This single command:

• Creates a local SQLite database at ~/.wisegit/wisegit.db
• Indexes your entire git history (462 commits in ~13 seconds)
• Creates .mcp.json for Claude Code auto-discovery
• Creates CLAUDE.md rules that instruct AI to check before editing
• Adds .mcp.json to .gitignore

2. Enrich with Issue Context (optional)

# Fetch issue/PR details from GitHub/GitLabGITHUB_TOKEN=ghp_... npx @sandip124/wisegit enrich

This fetches referenced issues (e.g., #134 in commit messages), detects Won't Fix / By Design decisions, and boosts freeze scores for functions linked to those issues.

3. Done

Open the repo in Claude Code. It will automatically:

1. Start the wisegit MCP server (via .mcp.json)
2. Read the protection rules (via CLAUDE.md)
3. Call get_file_decisions before editing any file

MCP Tools

MCP Resource: wisegit://manifest/{filePath} — decision manifest as auto-discoverable resource

MCP Prompt: check_before_edit — mandatory workflow prompt that returns the decision manifest before editing any file

LLM Intent Extraction Strategy

wisegit uses a smart fallback chain for extracting intent from NOISE commits:

Inside Claude Code, call extract_intent to retroactively recover intent for NOISE commits — uses Claude itself, no Ollama installation needed.

CLI Commands

wisegit setup [--path <dir>] [--global]         # One-command repo setupwisegit init [--full-history] [--path <dir>]     # Index git historywisegit enrich [--path <dir>]                    # Fetch issue/PR context from GitHub/GitLabwisegit audit <file>                             # Show decision manifestwisegit history <target> [--file <path>]         # Show decision timelinewisegit recompute [--path <dir>]                  # Recompute scores with PageRank + theory gapswisegit override <fn> --file <f> --reason "..."  # Override a frozen functionwisegit overrides                                # List active overrideswisegit sync                                     # Rebuild local cache from git + .wisegit/wisegit config list                              # View team configurationwisegit config set <key> <value>                 # Modify team policywisegit team-status                              # Team overview: enrichments, overrides, contributorswisegit team-health                              # Theory health: healthy/fragile/critical functionswisegit branch-capture                           # Capture branch context from last mergewisegit branch-list                              # List all captured branch snapshotswisegit branch-recover <sha>                     # Recover context from old merge commitwisegit calibrate                                # Show adaptive obsolescence weights vs defaultswisegit report [--output <file>]                 # Generate HTML report with scores + insightswisegit serve                                    # Start MCP server (stdio)wisegit hook install|uninstall                   # Manage git hooks (post-commit + post-merge)

Configure for Claude Code

Option A: Per-repo (recommended)

Run npx @sandip124/wisegit setup in any repo. It creates .mcp.json automatically.

Option B: Global registration

claude mcp add wisegit -- npx @sandip124/wisegit serve

Option C: Manual .mcp.json

Create .mcp.json in your repo root:

{  "wisegit": {    "command": "npx",    "args": ["@sandip124/wisegit", "serve"]  }}

Supported Languages

More languages can be added via Tree-sitter grammar configs in src/ast/languages/.

Issue Enrichment

A commit saying fix: handle null token #134 points to an issue containing reproduction steps, root cause, and explicit decision rationale — everything the commit message never says.

# Fetch issue context from GitHub/GitLabwisegit enrich --path /path/to/repo
# With auth (5000 req/hr instead of 60)GITHUB_TOKEN=ghp_... wisegit enrich

Supported platforms: GitHub, GitLab (Azure DevOps, Jira, Bitbucket planned)

Auth tokens: GITHUB_TOKEN / GH_TOKEN for GitHub, GITLAB_TOKEN for GitLab. Never stored by wisegit.

Issue-derived freeze signals

Freeze Score Signals

The freeze score is never stored directly — it's derived by replaying the event stream for each function. Signal categories:

Freeze score formula:

freeze_score = base_score x (1 - obsolescence_penalty)

Protection signals produce the base score (weighted average of present signal categories, not additive sum); obsolescence signals produce the penalty. Obsolescence weights are adaptive — calibrated per-repository using Shannon entropy and Bayesian feedback. Falls back to hardcoded defaults when < 20 functions have signals. Age signal uses Weibull survival model [18] (k=0.7, lambda=2.4y); contributor expertise uses DOE model [19] (4 variables: contribution share, recency, duration, frequency).

Academic grounding: 24 published papers. See REFERENCE.md for full citations.

Cross-Repo Validation

Tested on 3 real-world open-source codebases:

Flask's core APIs (__init__, run, wsgi_app, url_for) correctly scored FROZEN (0.87+). Express routing primitives (paramCallback, Route, Router) correctly scored FROZEN/STABLE. Scores adapt to each codebase's history rather than producing uniform distributions.

See REFERENCE.md for detailed validation findings and implementation changes.

Legacy Codebase Evolution

wisegit is designed for codebases that have accumulated years of intentional decisions. The freeze score doesn't mean "never change this" — it means "understand these decisions before you change it."

Progressive migration, not shiny rewrites. Per Távora [12]: the business rules in messy code are correct and valuable. The technical debt is in the structure, not the decisions. wisegit protects the decisions while you fix the structure.

See REFERENCE.md for the full legacy evolution section with academic grounding (24 published papers).

Team Support

wisegit uses a three-layer architecture — no separate "team mode" needed:

.wisegit/                      # Tracked by git — shared with team├── config.json                # Team policy (thresholds, AI authors)├── enrichments.jsonl          # Issue enrichment cache├── overrides.jsonl            # Override audit trail└── branch-contexts.jsonl      # Branch merge snapshots

JSONL format — one JSON object per line. Concurrent appends produce no git merge conflicts.

After a teammate pushes .wisegit/ changes, run wisegit sync to import them into your local cache.

See TEAM-ROADMAP.md for the full team architecture design.

Architecture

┌─────────────────────────────────────────────────┐│  Claude Code / MCP Client                       ││  ┌───────────────────────────────────────────┐  ││  │ 1. Reads CLAUDE.md protection rules       │  ││  │ 2. Calls get_file_decisions before edits  │  ││  │ 3. Respects FROZEN / STABLE / OPEN        │  ││  └───────────────────────────────────────────┘  │└─────────────────┬───────────────────────────────┘                  │ MCP (stdio)┌─────────────────▼───────────────────────────────┐│  wisegit MCP Server                             ││  ┌──────────┐ ┌──────────┐ ┌─────────────────┐ ││  │get_file_ │ │get_freeze│ │search_decisions │ ││  │decisions │ │_score    │ │                 │ ││  └────┬─────┘ └────┬─────┘ └───────┬─────────┘ │└───────┼─────────────┼───────────────┼───────────┘        │             │               │┌───────▼─────────────▼───────────────▼───────────┐│  SQLite (~/.wisegit/wisegit.db)                 ││  ┌──────────────┐ ┌────────────┐ ┌───────────┐ ││  │decision_events│ │freeze_scores│ │issue_     │ ││  │(append-only) │ │(derived)   │ │enrichments│ ││  └──────────────┘ └────────────┘ └─────┬─────┘ │└────────────────────────────────────────┼────────┘                                         │┌────────────────────────────────────────▼────────┐│  Issue Enrichment (wisegit enrich)              ││  ┌─────────┐ ┌─────────┐ ┌──────────────────┐  ││  │ GitHub  │ │ GitLab  │ │ Jira (planned)   │  ││  │ REST API│ │ REST API│ │                  │  ││  └─────────┘ └─────────┘ └──────────────────┘  │└─────────────────────────────────────────────────┘

Environment Variables

Security

• Everything runs locally — only issue enrichment makes outbound API calls (opt-in via wisegit enrich)
• Append-only event store — decisions are never deleted, only added
• SQLite database stored at ~/.wisegit/wisegit.db — no network exposure
• MCP tool inputs validated with strict Zod schemas (path traversal protection, length limits)
• Error messages sanitized before returning to MCP clients
• File writes check for symlinks before writing
• Config files parsed with allowlisted keys only (no prototype pollution)

Roadmap

• Phase 1 — Event store, AST chunking, commit classification, intent extraction, MCP server, CLI
• Phase 1.5 — Issue enrichment (GitHub, GitLab) with Won't Fix/By Design detection, freeze boost signals
• Phase 2 — Full freeze score: call graph + PageRank, theory gap detection (Naur death, forgotten patterns), co-change signals, Aranda signals, Ollama client, Go + Rust support
• Phase 4 — Override system (mandatory reason, time-boxed expiry, audit trail), branch context preservation (post-merge hook, snapshot storage, recovery)
• Phase A — Shared team knowledge layer: .wisegit/ directory with JSONL files for enrichments, overrides, branch contexts, and team config
• Phase B — Team-aware manifests: theory holder tracking, risk levels (healthy/fragile/critical), team status + health commands
• Phase C — AI-era adaptations: commit origin detection (HUMAN/AI_REVIEWED/AI_UNREVIEWED), origin-weighted freeze scores
• Phase D — Override approval workflow, team health metrics
• Phase E — Adaptive obsolescence calibration: 8 obsolescence signals, entropy-calibrated weights, Bayesian feedback, wisegit calibrate CLI

License

MIT