Using Claude Code straight out of the box means using maybe half of what you have. These five plugins each solve one thing: access to free model pools, memory across sessions, token savings, setup recommendations tailored to your project, and a system that improves your skills on its own. Below is what each one does and how to install it.
Claude CodeEklentiGeliştirici Araçları
01
OmniRoute: free AI gateway
Aggregates more than 350 providers behind a single endpoint. You point tools like Claude Code, Cursor, or Codex at it; when one provider hits its limit, it automatically falls back to another. So you don't stop mid-task because of a rate limit.
Installation
npm install -g omniroute
omniroute
# Dashboard: http://localhost:20128
# Before connecting a provider, check its terms flag
# on the Free Tiers page and skip anything marked avoid.
# For Claude Code, in a second terminal:
omniroute run claude --model auto
# For tools like Cursor, Codex, Cline:
# Base URL: http://localhost:20128/v1
# API Key: copy from Dashboard → Endpoints
# Model: auto
Needs Node 22.22.2 or above, or 24 through 26. When connecting Claude Code by hand, do not add /v1 to the address; that is only for other tools. Kiro AI and OpenCode Free, which the README suggests, are flagged avoid in the project's own terms of service table. The 'npm warn ERESOLVE' warnings during install are harmless. The detailed setup with every pitfall is at number 115.
Claude Code meets you from scratch every session; this plugin fixes that. It compresses and stores context from previous sessions and restores it automatically when you open a new one. On long-running projects it saves you from explaining everything again each time.
Installation
npx claude-mem install
# Or from inside Claude Code:
/plugin marketplace add thedotmack/claude-mem
/plugin install claude-mem
# Restart Claude Code after installing.
'npm install -g claude-mem' does NOT install it; that only pulls the library, it doesn't register the hooks. Use one of the two methods above.
Compresses everything your agent reads (tool outputs, logs, files, conversation history) before it reaches the model. Around 60-95% savings on JSON data and 15-20% on coding agents. Same answers, smaller bill.
Installation
# Python tool (not Node)
uv tool install --python 3.13 "headroom-ai[all]"
# or: pip install "headroom-ai[all]"
headroom wrap claude # wrap Claude Code
headroom doctor # verify the setup
headroom dashboard # watch the savings live
This is a Python tool, not Node. 'npm install headroom-ai' only gives you the TypeScript SDK; it does not ship the 'headroom' CLI.
Scans your project and recommends the hooks, skills, MCP servers, subagents, and slash commands that fit it. Read-only; it never modifies your files. The fastest path to a correct setup if you're new to Claude Code.
Installation
# Inside Claude Code:
/plugin install claude-code-setup@claude-plugins-official
# Then just ask:
"recommend automations for this project"
Task Observer: the meta-skill that improves your skills
It watches in the background while you work: the corrections you make, the tasks you repeat, and the gaps no skill covers yet. Then it proposes new skills and updates existing ones. It observes itself too. Its author reports applying 900+ improvements across 50 skills in the first six months.
Installation
# Download the repo as ZIP (Code → Download ZIP) or clone it.
# Keep SKILL.md and the references/ folder TOGETHER.
# For Claude Code:
.claude/skills/task-observer/
├── SKILL.md
└── references/
If you only drop in SKILL.md it runs degraded; the references/ folder is required.
Installing these tools on your own is one thing; actually building with them is another. The community has people using these daily and people building systems from scratch.
DOA: Yapay Zeka ve Otomasyon
If you want a system that actually runs in your business, let's talk for 10 minutes; I'll look at what you're trying to build and tell you which path fits. Free, and not a sales pitch.