100monkeys.ai

# 100monkeys.ai

> 100monkeys.ai builds AEGIS, a self-hosted orchestrator that runs autonomous agents in isolated, policy-enforced runtimes with durable FSM state, and four products alongside it: Carbon Forge, Your Zaru, Zaru CLI and Grill My Idea. AEGIS is open source under AGPL-3.0 and installs with one command: `curl -fsSL https://get.100monkeys.ai | bash`

AEGIS bounds what an agent can do at the infrastructure layer rather than in the prompt. Generate, execute, evaluate, refine: the loop runs until deterministic validators pass. The membrane around it enforces isolation and policy on every tool call, and each run is a durable state machine.

## AEGIS

- [100monkeys.ai](https://100monkeys.ai/): the company site, with the AEGIS overview and the install command
- [AEGIS documentation](https://docs.100monkeys.ai): the platform docs
- [Getting started](https://docs.100monkeys.ai/docs/getting-started): install and run a first agent
- [AEGIS documentation for LLMs](https://docs.100monkeys.ai/llms.txt): the docs index in llms.txt form
- [aegis-orchestrator on GitHub](https://github.com/100monkeys-ai/aegis-orchestrator): the source, AGPL-3.0

## Products

- [Carbon Forge](https://carbonforge.dev/): code, files and prose in one record on the server, which people and agents read and write through the same tools over MCP. Public workspaces are free, and every account includes one free private workspace
- [Your Zaru](https://myzaru.com/): a personal AI companion that works alongside you, remembers you and shows its work
- [Zaru CLI](https://github.com/100monkeys-ai/zaru-cli): a terminal coding agent that keeps its execution loop in view, open source under Apache-2.0
- [Grill My Idea](https://grillmyidea.com/): market research for a startup idea, before you build it

## Read

- [FAQ](https://100monkeys.ai/faq): short answers on self-hosting an agent orchestrator, keeping credentials away from agents, durable agent workflows and accountability
- [Red levers: traceable, deterministic agents in production](https://100monkeys.ai/papers/red-levers): a paper by Jeshua ben Joseph, 100monkeys AI, Inc., CC BY 4.0 ([PDF](https://100monkeys.ai/papers/red-levers.pdf))
- [Datasheets](https://100monkeys.ai/datasheets/): one-page PDFs for AEGIS, Carbon Forge and Zaru
- [Blog](https://100monkeys.ai/blog/): essays from 100monkeys.ai ([RSS](https://100monkeys.ai/blog/rss.xml))
- [Blame that remembers why](https://100monkeys.ai/blog/blame-that-remembers-why): what a change should carry when the author was an agent
- [The approval is the product](https://100monkeys.ai/blog/the-approval-is-the-product): why an agent orchestrator should make you say yes
- [Run a scheduled agent with AEGIS](https://100monkeys.ai/blog/run-a-scheduled-agent-with-aegis): Put an AEGIS agent on a timer, fire it once by hand, and read each run: started, skipped, refused or paused. Every field from the public docs.
- [Reading an AEGIS execution log and a sealed tool call](https://100monkeys.ai/blog/reading-an-aegis-execution-log): Read an AEGIS run from its execution record to its iteration log to the SEAL envelope and audit event behind each tool call, allowed or refused.
- [AEGIS on Firecracker: what the docs describe today](https://100monkeys.ai/blog/aegis-on-firecracker): Firecracker support in AEGIS is planned, not shipped. Check a host for KVM, read the documented config, and run the Podman path that works now.
- [Install AEGIS on Ubuntu and run an approved task](https://100monkeys.ai/blog/install-aegis-on-ubuntu-vps): Install AEGIS on a fresh Ubuntu server with one command, run an agent, then gate its result behind a human approval. Every step from the public docs.
- [Instala AEGIS en Ubuntu y ejecuta una tarea aprobada](https://100monkeys.ai/blog/es/install-aegis-on-ubuntu-vps) (es): Instala AEGIS en un servidor Ubuntu con un solo comando, ejecuta un agente y deja su resultado tras una aprobación humana. Todo, de la documentación.
- [Instale o AEGIS no Ubuntu e rode uma tarefa aprovada](https://100monkeys.ai/blog/pt/install-aegis-on-ubuntu-vps) (pt): Instale o AEGIS num servidor Ubuntu com um comando, rode um agente e deixe o resultado atrás de uma aprovação humana. Cada passo vem da documentação.
- [Ubuntu に AEGIS を入れて承認付きタスクを実行する](https://100monkeys.ai/blog/ja/install-aegis-on-ubuntu-vps) (ja): 新しい Ubuntu サーバーに AEGIS をコマンド1つでインストールし、エージェントを動かして、その結果を人間の承認の後ろに置きます。手順はすべて公開ドキュメントから。
- [A traceability checklist for AI agents in production](https://100monkeys.ai/blog/ai-agent-traceability-checklist): What to log on every agent run, what to sign so the log can be trusted, and what to show the reviewer. Vendor-neutral, with open standards for each step.
- [Lista de trazabilidad para agentes de IA en producción](https://100monkeys.ai/blog/es/ai-agent-traceability-checklist) (es): Qué registrar en cada ejecución de un agente, qué firmar para que el registro sea fiable y qué mostrar a quien revisa. Neutral, con estándares abiertos.
- [Checklist de rastreabilidade para agentes de IA](https://100monkeys.ai/blog/pt/ai-agent-traceability-checklist) (pt): O que registrar em cada execução de um agente, o que assinar para o registro ser confiável e o que mostrar a quem revisa. Neutro, com padrões abertos.
- [本番 AI エージェントのトレーサビリティ・チェックリスト](https://100monkeys.ai/blog/ja/ai-agent-traceability-checklist) (ja): エージェントの各実行で何を記録し、記録を信頼できるものにするために何に署名し、レビュー担当者に何を見せるか。特定ベンダーに依存せず、各項目にオープン標準を添えています。
- [How to add an approval step to an agent orchestrator](https://100monkeys.ai/blog/approval-step-for-an-agent-orchestrator): A pattern for human approval gates in agent workflows: where the gate goes, what the reviewer sees, what silence means, with AEGIS as the worked example.

## Company

- [GitHub](https://github.com/100monkeys-ai): the 100monkeys-ai organization
- [LinkedIn](https://www.linkedin.com/company/112423436/): 100monkeys.ai
- [X](https://x.com/100monkeysai): @100monkeysai
- [Hashnode](https://100monkeys-ai.hashnode.dev): engineering articles