>DRY.codes
Sign in

DRY.codes turns your repositories into a knowledge base your coding agent can search. Before it writes something new, it finds what your team already built and reuses that instead.

Get started with GitHub See how it works

Try it on the public corpus ↓

Leaderboard

The 25 driest public repos

Public repositories shared on DRY.codes, ranked by originality. Internal is how little a repo repeats itself. Public is how original it is versus every other public repo. A wide gap between the two means the repo overlaps other public code, usually because it was forked, templated, or generated. Click a column to re-rank, or hover a score for the breakdown.

Showing the top 25 of 58 public repositories, ranked by Public DRY (originality vs the whole corpus).

#Repository Internal Public Files Indexed
🥇 zhw2590582/ArtPlayer
91
81
135 2026-07-16
🥈 mshumer/Claude-of-Duty
84
74
182 2026-09-12
🥉 mukul975/Anthropic-Cybersecurity-Skills
88
68
120 2026-06-27
4 quotentiroler/aegis
90
68
39 2026-09-07
5 Jakubantalik/border-beam
92
65
26 2026-07-27
6 every-app/open-seo
91
64
143 2026-06-26
7 nousresearch/hermes-agent
87
64
141 2026-06-25
8 mauriceboe/TREK
85
62
140 2026-06-26
9 koala73/worldmonitor
87
61
134 2026-06-25
10 alexheretic/ab-av1
85
61
31 2026-09-12
11 quotentiroler/CactusRoute
89
61
12 2026-09-07
12 andrewyng/openworker
79
60
383 2026-07-27
13 opendatalab/MinerU
79
60
145 2026-06-26
14 tamnguyenvan/screenarc
84
60
140 2026-09-12
15 Kaliiiiiiiiii-Vinyzu/patchright
86
60
63 2026-09-11
16 n0-computer/iroh
82
59
141 2026-06-27
17 aws/agent-toolkit-for-aws
80
58
145 2026-06-26
18 google-research/timesfm
79
58
81 2026-06-27
19 modelcontextprotocol/typescript-sdk
74
57
331 2026-09-09
20 AgriciDaniel/claude-seo
77
57
201 2026-09-10
21 IceWhaleTech/CasaOS
86
57
141 2026-06-27
22 comfy-org/ComfyUI-Manager
85
57
91 2026-09-12
23 shanraisshan/claude-code-best-practice
80
57
88 2026-06-26
24 interviewstreet/hiring-agent
93
57
14 2026-06-27
25 microsoft/vscode
87
56
139 2026-06-25

The problem

Duplication scales with your company

Every engineer you add writes more code, and some of it already exists somewhere in your repositories. Nobody does this deliberately. It happens because finding the earlier version is usually harder than writing a fresh one.

Coding agents raise the stakes. They write more code, faster, and they begin each task with no memory of what your team has already built. One engineer can now produce what used to take a whole group, and the duplication grows right along with it.

Duplicated code → engineers + coding agents 1 dev 10 100 1,000+ devs
Illustrative. Duplication already grew faster than the teams producing it, and coding agents lift the ceiling further.

Why DRY.codes

Save up to 80% in token usage

Every time an agent regenerates code that already exists, you pay twice: once for the tokens that write it, then again for the work of keeping two copies in step. DRY.codes gives your agent the memory to find what is already in your repositories, so it writes less and ships more.

Regenerate from scratch 80% wasted re-writing existing code Find & reuse with DRY.codes ~20% −80% tokens saved
Illustrative. The red stripes are tokens spent rewriting code that already exists, which is the part DRY.codes lets your agent skip.

The knowledge base

A knowledge base for everything your team has built

Indexing a repository adds its code to a connected map of how things are already built. Every file becomes a node, and every shared pattern becomes a link. Your agent queries that map over MCP to find what exists across your repositories, and across the public corpus, before it writes a new line. The more you index, the more it knows.

repo-a repo-b repo-c DRY .codes reuse duplicate
Illustrative. Once your repositories are indexed they stop being islands. DRY.codes links the same function wherever it lives, so your agent can reuse it (green) rather than write another copy (amber).

How it works

From repo to reuse

  1. 1

    Connect your repos

    Sign in with GitHub and pick the repositories DRY.codes should index.

  2. ★

    Installrecommended

    The skill makes your agent check DRY.codes before it writes, turning the rule at the top of this page into something it follows every time.

    One command installs the skill and connects DRY.codes. Sign in, then pick your repositories or the public corpus:

    /plugin marketplace add max-network/plugins
/plugin install dry-codes@max-network

    Register the marketplace, then install the dry-codes plugin (or enable it from Codex's plugin directory):

    codex plugin marketplace add max-network/plugins
codex plugin add dry-codes@max-network

    Point any MCP client at https://dry.codes/mcp and sign in, then pick an endpoint or the public corpus. If your client does not support OAuth, use a per-endpoint URL from the dashboard. Then install the skill:

    curl -sL --create-dirs \
  -o .claude/skills/dry-codes/SKILL.md \
  https://dry.codes/skill/SKILL.md

    View the skill · works with Claude Code, Codex, Claude.ai, and the Agent Skills standard.

  3. 2

    Point your agent at it

    Connect https://dry.codes/mcp to Claude, Cursor, or any MCP client and sign in.

  4. 3

    Reuse what already exists

    Your agent finds and reuses what your team has already built, so you spend fewer tokens and your codebase stays coherent.

Start reusing what you already have

Index your first repository for free, connect your agent, and see what it finds. No credit card required.

Get started with GitHub View plans

Want to look around first? Search every public repository right here, no account needed.

FAQ

Questions people ask

What is DRY.codes?
DRY.codes indexes your GitHub repositories and makes them searchable by your coding agent over MCP, so it reuses code your team already wrote instead of writing it again. It also reports where your codebase repeats itself.
Does DRY.codes store my source code?
No. Indexing stores a derived index of your repositories, not a copy of your files. When a tool shows you the contents of a file, it is fetched from GitHub at that moment using your own installation's access.
What is MCP, and do I need it?
MCP (Model Context Protocol) is how AI coding tools connect to external tools and data. If you use Claude Code, Codex or another MCP client, you point it at your DRY.codes endpoint and your agent gains the search and duplication tools. You can also use the web dashboard on its own.
What is the DRY score?
A 0–100 rating of how little a codebase repeats itself, where higher is better. It combines three views — repeated sequences, repeated lines and repeated files — and the DRY Report lets you weight them yourself to see which kind of repetition dominates.
Do I have to give access to all my repositories?
No. You choose which repositories to index, and each MCP endpoint you create is scoped to the owners and repositories you pick — so an agent only ever sees the slice you pointed it at.
Can I use it on private repositories?
Yes. Private repositories stay private: they are never added to the public corpus, and nothing about them is shared with other accounts. Listing a repository publicly is an explicit, separate action, and only public repositories can be listed.
What does it cost?
Indexing your first repository is free, with no credit card. Paid plans raise the limits and add the semantic (meaning-based) tools — see pricing.
How is this different from a linter or a code-duplication checker?
A linter checks the file in front of you. DRY.codes answers a different question — does this already exist somewhere in our code? — across every repository you have indexed, and answers it to your agent while it is writing, not to you afterwards in a report.