bleuropa

bleuropa / loom

Public

An Elixir-native AI coding assistant with persistent decision graphs, repo intelligence, and a Phoenix LiveView UI

107
11
100% credibility
Found Feb 28, 2026 at 51 stars 2x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Elixir
AI Summary

Loom is an AI coding assistant that analyzes codebases, suggests and applies edits via chat, runs commands, and visualizes decision-making in a web interface or command line.

How It Works

1
🦉 Discover Loom

You hear about Loom, a friendly AI helper that understands your code and makes fixing bugs or adding features feel easy.

2
💻 Set it up

You download Loom and launch it with a simple click, watching it prepare to assist.

3
🤖 Connect smart AI

You pick a clever AI brain like Claude so Loom can think deeply about your code.

4
📁 Open your project

You point Loom to your coding folder, and it scans everything to get familiar.

5
💬 Chat about your code

You type questions like 'fix this bug' in the chat, and Loom responds with ideas.

6
Review and apply changes

Loom shows previews of edits, file differences, and a map of its thinking—you approve and it updates your code safely.

🚀 Your code improves

Your project runs better with tracked decisions and clean changes, ready for more work anytime.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 51 to 107 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 loom?

Loom is an Elixir-built AI coding assistant that analyzes your repo, suggests and applies code edits, runs shell commands, and handles git commits—all via a slick Phoenix LiveView web UI or interactive CLI. It tracks reasoning in a persistent decision graph stored in SQLite, so it remembers past goals, options, and outcomes across sessions, solving the "amnesiac AI" problem in tools like GitHub Copilot alternatives. Developers get repo intelligence like file trees, diffs, LSP diagnostics, and multi-provider LLM support (Anthropic, OpenAI, Groq) out of the box.

Why is it gaining traction?

Unlike Python/JS-heavy agents, Loom leverages Elixir's BEAM for crash-proof concurrency—tools run in parallel without GIL headaches—and delivers a real-time web UI with streaming chat, no custom JS needed. The decision graph and emerging agent teams (researcher/coder/reviewer roles coordinating via PubSub) provide structured memory that cheap models like GLM turn into collective smarts, making it a fresh GitHub loom alternative for architectury workflows. Early adopters dig the single-binary packaging and cost dashboard for tracking LLM spend.

Who should use this?

Elixir/Phoenix devs debugging tests or refactoring OTP apps will love the native integration and LiveView previews. Backend teams building AI agents (think dbt loom github for data pipelines or fabric loom github for infra) get fault-tolerant multi-agent swarms without external queues. Solo coders seeking a local, persistent coding assistant beyond Cursor or Aider.

Verdict

Try Loom if you're in the Elixir ecosystem—its 335 tests and polished UI scream potential despite 38 stars and 1.0% credibility score signaling early days. Maturity lags (Phase 5 agent teams in dev), but docs and binary builds make it dead simple to eval now.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.