🔥 Simplicio — The AI Agent That SAVES UP TO 96% OF YOUR TOKENS

Install · Features · 96% Savings · Website

🌍 Languages:
🇬🇧 English | 🇧🇷 Português | 🇪🇸 Español | 🇫🇷 Français | 🇯🇵 日本語 | 🇰🇷 한국어 | 🇨🇳 简体中文 | 🇮🇹 Italiano | 🇷🇺 Русский | 🇵🇱 Polski | 🇮🇳 हिन्दी | 🇸🇦 العربية | 🇮🇱 עברית | 🇲🇾 Bahasa Melayu | 🇮🇩 Bahasa Indonesia

---

⚡ TL;DR

Simplicio is a terminal AI coding agent — a single binary that replaces your entire AI-assisted development workflow: chat, code generation, repository context, planning, local multi-agent orchestration (64 → 600 agents), and evidence-backed PR delivery.

Runs on your machine. Your code never leaves your control. Remote models are optional, not required.

🔥 Save up to 96% of tokens vs traditional agents — more than Caveman (65%) or RTK (80%). Every interaction shows exactly how many tokens you saved. Single Rust binary, zero deps.

🚀 Installation

macOS / Linux

curl -fsSL https://raw.githubusercontent.com/wesleysimplicio/simplicio/master/install.sh | sh

Windows

powershell -c "irm https://raw.githubusercontent.com/wesleysimplicio/simplicio/master/install.ps1 | iex"

Done. One command downloads the Runtime binary, verifies its checksum, and validates the embedded Python source bundle. The installer does not use a Python package manager or sibling checkout. A Google login with an active entitlement is mandatory before product commands run; public beta access may be free, but it does not bypass login.

Codex MCP authentication

After login, the installer registers Simplicio as a local Streamable HTTP MCP server at http://127.0.0.1:8787/mcp. Open Codex MCP settings: the Simplicio row will show Authenticate. Click it to use the Google-backed login at simpleti.com.br/simplicio/login. The Runtime validates the bearer token and active entitlement on each MCP request; beta access does not bypass this gate.

Other MCP hosts keep the local stdio fallback. If the host was configured before this release, run simplicio mcp register and reload its MCP settings.

Python projects embedded in the Runtime

The Runtime release is the distribution boundary. The generated binary carries the real Python source trees for the six projects below. They remain Python projects; the Runtime host is not a native-Rust rewrite and normal installation does not clone or download their repositories:

Capability	Binary surface
Mapper	Embedded Mapper Python source + Runtime bridge
Dev CLI	Embedded Dev CLI Python source + Runtime bridge
Loop	Embedded Loop Python source + Runtime bridge
Fast	Embedded Fast Python source + Runtime bridge
Prompt	Embedded Prompt Python source + Runtime bridge
Sprint	Embedded Sprint Python source + Runtime bridge

Verify a release after installation:

simplicio ecosystem verify --json

The verification output includes the source archive size/digest, embedded component versions, provenance commits, and compatibility status. Updating the Runtime release updates this bundle. The binary still uses an available Python 3 interpreter to execute the embedded sources; it never downloads packages or falls back to sibling checkouts. Legacy external adapters and portable Claude skills may still be used explicitly, but they are outside the normal path.

After installation, authenticate with the Google-backed device flow when prompted, or run:

simplicio auth login
simplicio auth status --json

Plugin marketplace

This public repository also publishes the Simplicio Claude Code marketplace:

/plugin marketplace add wesleysimplicio/simplicio
/plugin install simplicio-loop@simplicio
/plugin install simplicio-prompt@simplicio
/plugin install simplicio-sprint@simplicio

The plugin bundle is documented in PLUGIN.md. These are optional Claude Code skill surfaces; they are not required to obtain the Runtime's embedded ecosystem. When present, the skills call the Runtime through simplicio serve --mcp --stdio.

---

💰 Token Savings — 96% is Real

Without Simplicio: every AI session rediscovers your repo, loads too much context, repeats prompts, burns paid tokens.

With Simplicio:

Optimization	Savings
🗺️ Repo Map — compressed context instead of reading raw files	~70%
🧠 Memory Recall — known facts are not re-derived	~80%
✏️ Deterministic Editing — changes without spending LLM tokens	100% (output)
🏠 Local LLM — classification, summarization, low-risk edits	~90%
📡 Remote LLM — only for planning and complex decisions	~85%
🔀 Local Fan-out — 64→600 agents before scaling to cloud	~95%
💎 Combined: up to 96% total savings	~96%

Every Simplicio response shows real savings: Simplicio: ~X tokens spent · saved ~Y (Z%)

---

🎯 What It Does

Command	Description	Tokens
simplicio map --repo .	Maps the repository for LLMs	~70% savings
simplicio memory "query"	Neural recall (FTS + vectors)	~80% savings
simplicio edit '{...}'	Deterministic file editing	Zero tokens
simplicio coding-loop "task"	Iterates until tests pass	Auto-repair
simplicio deliver certify	5 quality gates before shipping	Deterministic
simplicio run "task" --agents N	Multi-agent orchestration	Local-first

---

🆚 Simplicio vs Caveman vs RTK

	🪨 Caveman	🔧 RTK	🔥 Simplicio
Approach	Output style compression	Shell command proxy	Full agent runtime
Max savings	~65% output tokens	~80% on shell commands	Up to 96% total
Input compression	❌	✅ (filtered)	✅ Repo map + neural memory
Output compression	✅ (caveman-speak)	❌	✅ Zero-token deterministic edits
Local LLM	❌	❌	✅ Built-in llama.cpp
Multi-agent	❌	❌	✅ 64 → 600 local agents
Memory across sessions	❌	❌	✅ FTS + vector recall
Evidence chain	❌	❌	✅ sha256 sealed receipts
Language	JS/Python (skill)	Rust (binary)	Rust (single binary)
License	MIT	Apache 2.0	Proprietary
Stars	72.5k	62.2k	⭐ You're early

Bottom line: Caveman makes the AI talk less. RTK makes commands output less. Simplicio makes the AI think less — by remembering, mapping, editing deterministically, and running locally before ever touching a paid LLM.

| Simplicio saves 96% where Caveman saves 65% and RTK saves 80%. |

---

🏗️ Architecture

LLM (Claude/Codex/Gemini)          Simplicio Runtime (Rust)
  |                                   |
  | 1. Orient                         | simplicio map
  | 2. Recall                         | simplicio memory
  | 3. Decide                         |
  | 4. Edit  ───────────────────────> | simplicio edit (0 tokens)
  | 5. Verify <─────────────────────  | simplicio deliver certify
  | 6. Iterate                        | simplicio coding-loop

The LLM reasons. Simplicio executes deterministically.

---

✨ Features

• 🏠 Local-first — built-in llama.cpp, scales to remote only when needed
• 🪜 Tiered agents — 64 → 100 → 200 → 600 local agents before paid cloud
• 🔇 Shannon novelty gate — filters redundant outputs (zero tokens on dedup)
• 🔒 Sealed receipts — sha256 per artifact, tamper-proof evidence chain
• 🛡️ 5 delivery gates — acceptance, validation, run-verify, regression, self-review
• ⚡ Action gate — risk classification + blocklist for chat-initiated mutations
• 🔌 MCP/ACP — Model Context Protocol + Agent Client Protocol
• 🌐 Gateways — Telegram, Discord, Slack, WhatsApp
• 🧩 Skill system — loads and chains reusable capabilities
• 💾 Memory DB — persistent FTS + vector recall across sessions
• 🔀 LLM router — no LLM → local LLM → remote LLM automatically
• 🖥️ Cross-platform — macOS, Linux, Windows, single binary

---

🎁 Public Beta

Deterministic commands are FREE forever: map, validate, edit, deliver, checkpoint

AI features may be free while the public beta flag is active. Login and an active entitlement are still required; when beta access ends, the entitlement must come from an active subscription.

simplicio license status

---

📋 Requirements

Requirement	Minimum	Recommended
RAM	8 GB	16 GB+
Storage	5 MB	1.5 GB (with local LLM)
OS	macOS 13+, Linux, Windows 10+	macOS ARM64
Terminal	any modern terminal	WezTerm / Alacritty / Ghostty

---

🧪 Testing this repo's tooling

This repo ships committed release binaries plus the packaging/tooling around them (npm/PyPI wrappers, install scripts, a distribution-consistency checker). The official command to run that tooling's unit test suite:

pip install -r requirements-dev.txt
python -m pytest tests/unit -v --cov=scripts --cov-report=term-missing --cov-fail-under=85

See docs/testing-strategy.md for what's covered, what's intentionally out of scope, and the plan for the rest of the testing epic. See also CONTRIBUTING.md.

---

🌐 Ecosystem

• Website — full docs, benchmarks, install
• Discord — community and support

---

📄 License

Proprietary. Binary free to download and use. AI features free during the public beta. See LICENSE.

---

⭐ Star History

---

💬 Community

• Discord — chat, support, early access
• GitHub Issues — bugs and feature requests

---

🔥 Simplicio — Your code, your machine, 96% cheaper. 🔥