sdwolf4103

sdwolf4103 / opencode-working-memory

Public

Four-tier memory architecture for OpenCode AI agents: persistent core memory, session working memory, smart pruning, and pressure monitoring

37
0
100% credibility
Found Feb 18, 2026 at 23 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A plugin for an AI coding assistant that adds persistent memory for goals, progress, errors, and context to maintain focus across long sessions.

How It Works

1
🔍 Discover the memory helper

You find this handy add-on that helps your AI coding friend remember important details like goals and mistakes so it stays sharp over long chats.

2
Pick your easy setup
🤖
Let AI handle it

Paste a quick note to your AI and it installs everything for you.

📝
Simple settings tweak

Add one line to your AI's settings file and restart.

3
🚀 Everything is ready

Your AI helper now has super memory built-in, working quietly in the background.

4
🧠 Set your first goal

Chat with your AI and say 'remember my goal is to fix this bug'—it saves it forever, even after resets.

5
💡 Watch it remember automatically

As you work on files or spot errors, your AI grabs and keeps the key facts without you lifting a finger.

6
⚠️ Smart full-memory alerts

When things get crowded, it warns you and suggests quick ways to tidy up and keep going smoothly.

✅ AI stays focused forever

Now your AI buddy never forgets your project details, saving time and frustration on big tasks.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 23 to 37 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 opencode-working-memory?

This TypeScript plugin equips OpenCode AI agents with a four-tier memory architecture: persistent core memory for goals and progress, session working memory for errors and decisions, smart pruning for tool outputs, and pressure monitoring for token limits. It tackles agents forgetting context after compactions by auto-preserving high-value info like todos, dependencies, and file paths across resets. Users get tools like core_memory_update and working_memory_add with zero configuration—just add to opencode.json and restart.

Why is it gaining traction?

It stands out with real-time pressure monitoring that triggers warnings at 75% and interventions at 90% usage, plus adaptive pruning that compresses outputs based on tool type and load. Developers see agents avoid repeating mistakes, stay focused in long sessions, and self-manage storage with TTLs and file limits. The hook is seamless continuity without token waste or disk sprawl.

Who should use this?

OpenCode users running extended sessions on complex TypeScript or multi-file projects, where agents track evolving goals, bugs, and file paths. Suited for solo devs or teams using AI for iterative coding, refactoring, or debugging who hit context limits often. Avoid if you restart sessions frequently for clean slates.

Verdict

Try it if you're deep into OpenCode agents—strong docs and instant value outweigh the 16 stars and 1.0% credibility score. Early maturity means test in non-critical workflows first.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.