
The AI coding agent for building ambitious software
memU records 7,207 monthly visits with -29.6% month-over-month growth and Freemium pricing. Repowise records 29,455 monthly visits with +26.3% growth and Freemium pricing.

One shared, inspectable memory layer for every AI agent you use
| Dimension / Metric | ![]() | ![]() |
|---|---|---|
| Official Domain | memu.pro | repowise.dev |
| Pricing Model | Freemium | Freemium |
| Monthly Visits | 7.21K (7,207) | 29.46K (29,455) |
| MoM Traffic Delta |
| -3.03K (-29.6%) |
| +6.14K (+26.3%) |
| Primary Market & Share | 🇺🇸 United States (48.3%) | 🇮🇳 India (20.6%) |
|---|
| Direct Visits / mo | 2.85K | 13.92K |
|---|
| User Rating | 4.3 | 4.6 |
|---|
| Supported Platforms | Website | Website |
|---|

One shared, inspectable memory layer for every AI agent you use
One memory store shared by Codex, Claude Code, Cursor, OpenClaw, Hermes, WorkBuddy and other agents, so a lesson learned in one tool is available in the next.
Memories are written as plain Markdown files with summaries, sources and related-note links, so you can read, edit or remove any record yourself.
memU watches sessions for durable patterns and preferences, turning repeated corrections into reusable skills and habits rather than one-off prompts.
Every retrieved context pack keeps links back to the original conversation or document, so you can verify why the agent believes something.

One living index of your codebase that AI agents and pull requests can actually read
Structure, call edges, git churn, co-change patterns, ownership, test reachability and recorded decisions are mapped once and refreshed on every commit, giving agents and reviewers a single shared source of truth.
Claude Code, Cursor, Codex, VS Code and any MCP client can ask for the callers, owners and tests of a file before editing it, which the vendor benchmarks at roughly a third fewer agent output tokens on large repos.
Every file receives a 1-to-10 score from deterministic markers validated against real bug fixes, with ranked refactoring plans, dead-code findings and cross-file performance issues such as N+1 queries.
One comment per PR names the callers a change reaches, lists changed lines that no test covers, stays silent when the PR is clean and adapts to replies your team writes.

The AI coding agent for building ambitious software
Google's agent-first development platform for building and shipping software with AI agents

Free / Open-sourceRun and build with open AI models locally or in the cloud
#3
The AI coding agent that turns ideas into shipped software
Free / Open-sourceA free, open-source personal AI agent that learns your projects and works across every messaging app.
Cognition's autonomous software engineer that ships real pull requests

The coding agent that learns your coding taste

Free / Open-sourceA private, memory-powered AI assistant you own and customize

Agent orchestration platform for your whole software development lifecycle

Free / Open-sourceCut AI agent token costs with recoverable context compression and smart model routing

Free / Open-sourceOpen-source workspace to build, deploy and govern AI agents

AI-native offensive security: continuous red teaming, white-box pentests, and PR review
The AI coding agent for building ambitious software
Google's agent-first development platform for building and shipping software with AI agents
Free / Open-sourceRun and build with open AI models locally or in the cloud
Free / Open-sourceA free, open-source personal AI agent that learns your projects and works across every messaging app.
The coding agent that learns your coding taste
Free / Open-sourceA private, memory-powered AI assistant you own and customize
Agent orchestration platform for your whole software development lifecycle
Free / Open-sourceCut AI agent token costs with recoverable context compression and smart model routing