ameya1995

ameya1995 / Constrictor

Public

An agent-first dependency and blast-radius explorer for Python codebases. Generates structured, machine-readable dependency graphs that AI coding agents can query to understand code relationships, assess refactoring impact, and navigate complex projects autonomously.

15
1
100% credibility
Found Mar 18, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Constrictor builds interactive dependency graphs from Python codebases to visualize call flows, imports, frameworks, and databases, helping predict the impact of changes.

How It Works

1
📰 Discover Constrictor

You hear about a helpful tool that maps out how all the pieces in your Python project connect, so you know what might break when making changes.

2
📥 Add to your project

You simply download and place the tool in your project's folder, ready to explore your code.

3
🔍 Build your code map

You run a quick scan, and it creates a complete picture of functions, classes, routes, and data flows across your entire codebase.

4
🌐 View the interactive graph

Open a browser window to see a colorful, zoomable map where you can filter views and spot services or data connections.

5
⚠️ Check change impacts

Before editing, select any piece of code to instantly see what depends on it or what it affects downstream.

✅ Refactor with confidence

You safely update your code, remove unused parts, and fix issues, knowing exactly the ripple effects thanks to your clear dependency map.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is Constrictor?

Constrictor scans Python codebases to build structured dependency graphs capturing imports, call graphs, class hierarchies, framework endpoints from FastAPI, Flask, Django, SQLAlchemy models, HTTP calls, and type annotations. It answers "what breaks if I change X?" via blast-radius analysis, letting developers assess refactoring impacts before touching code. Agent-first design outputs machine-readable JSON that AI coding agents query to navigate complex projects autonomously.

Why is it gaining traction?

Unlike basic static analyzers, Constrictor integrates directly with AI agents via MCP servers for Claude Code, Cursor, and Copilot, plus SKILL.md for skill-file discovery—enabling autonomous code exploration in agent workflows. CLI commands like `impact`, `diff-impact`, `paths`, `unused`, and `cycles` deliver instant insights, with a D3 web UI for visual topology and cross-service boundaries. Optional JS/TS support stitches frontend HTTP calls to backend endpoints, handling real-world stacks.

Who should use this?

Backend teams maintaining large Python services with FastAPI or Django, needing quick PR impact checks via `git diff | constrictor diff-impact`. AI agent users in Cursor or Claude Code who want agents to assess blast-radius before edits. Devs hunting dead code or cycles in monorepos with docker-compose multi-services.

Verdict

Early alpha (15 stars, 1.0% credibility) with solid docs and CLI, but not PyPI-published yet—install from GitHub source. Worth trying for agent-first blast-radius in complex codebases; pair with incremental `--watch` for live graphs.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.