CLOUDM0N
Repository intelligence

rush86999/atom

GitHub

An open-source, self-hosted AI agent workforce and governance platform that deploys specialized agent teammates to execute automated business workflows inside a deterministic safety net on local or cloud-based hardware.

CLOUDM0N decision
Trust FAIL · 0/100
Good fit if

Teams needing governed, multi-agent business automation with strict data-privacy requirements.

Watch out for

Requires user-managed infrastructure, which places the complete operational overhead of self-hosting and maintaining the embedded store and local network on the user.

Practical intelligence

What matters before you adopt it

Problem it solves

The extremely low success rate of bringing AI agent pilots into production due to security gaps, lack of reliable governance, unverified self-reporting, and privacy leakage.

Best for
Teams needing governed, multi-agent business automation with strict data-privacy requirements.
Organizations seeking to transition from traditional static step-based workflow tools (such as Zapier, Make, or n8n) to reasoning, self-correcting AI agents.
Users subject to EU AI Act data-governance obligations who require fully private, self-hosted LLM deployments.
Main trade-offs
Requires user-managed infrastructure, which places the complete operational overhead of self-hosting and maintaining the embedded store and local network on the user.
Why it stands out
Free, fully-featured, and open-source under the AGPL v3 license with no closed-source pro build tier.
Excellent latency with 0.027ms P99 per governance check and a cached throughput of 616k ops/s.
Deep security sweep history comprising 69+ documented TDD hardening rounds and 85k+ test functions.
Trust & CVEs

Security evidence without the noise

Trust remains a decision signal; CVEs and scanner evidence explain what is driving the risk.

Security findings
7
CLOUDM0N scanner findings
Critical
1
High
3
Medium
3
Low
0
View trust evidence & security findings▼
Why this score
No trust rationale was stored for this scan.
CLOUDM0N findings
CRITICAL
Private-key material appears to be committed in executable/config scope. 3 sample match(es) found.
HIGH
Remote download piped or chained into a shell requires manual review. 5 sample match(es) found.
HIGH
Container configuration requests host-level control or isolation bypass. 5 sample match(es) found.
HIGH
4 high severity dependency vulnerability/vulnerabilities reported by npm audit.
MEDIUM
Dynamic code execution pattern detected. 5 sample match(es) found.
MEDIUM
The project can spawn operating-system processes; review command construction and input handling. 5 sample match(es) found.
MEDIUM
Broad permission or elevated-command pattern detected. 5 sample match(es) found.
Architecture from code4 modules · 4 edges▼
Structural evidence

Modules and dependency edges extracted from repository code. This is code evidence, not README inference.

Code files
3504
Modules
4
Dependency edges
4
Core modules
(root)
1 files
archive
120 files
backend
3383 files
Dependency flow
(root)→backend
archive→(root)
archive→backend
backend→(root)
Detected languages
Python · Shell · TypeScript · JavaScript · SQL
Detected frameworks
Anthropic SDK · FastAPI · NumPy · OpenAI SDK · PyTorch · Pydantic · Requests · SQLAlchemy · aiohttp · pandas · scikit-learn
Architecture evidence details
flowchart TD
    %% atom-root — high-level architecture (DRAFT, refine me)
    n0["(root) · 1 file"]
    n1["archive · 120 files"]
    n2["backend · backend · 3383 files"]
    n0 --> n2
    n1 --> n0
    n1 --> n2
    n2 --> n0
    class n2 api
    classDef api fill:#af7aa1,color:#ffffff,stroke:#8a5f7f
Evidence, security & integrations▼
Integrations
SalesforceHubSpotSlackTeamsGmailNotionJiraLinear
Security notes
• Runs a default-on execution sandbox that restricts filesystems, tools, egress traffic, and resources at every tool-dispatch hub.
• Provides an outbound gatekeeper that enforces rate limits, response masking, and data-taint tracking to block sensitive leaks.
• Features robust access control including OIDC SSO, SCIM v2 provisioning, and an 8-role RBAC model.
• Encrypts OAuth integration tokens at rest using Fernet, and the production environment fails closed if keys are missing.
Still unknown
• The README does not provide minimum hardware specifications (such as CPU, RAM, or disk storage limits) required to host the database, FastAPI backend, and Next.js frontend.
• Does not detail the exact database engines used for the embedded transactional or vector stores besides general mentions of SQL, LanceDB, and SQLite.
• Lacks information regarding upgrade paths, database migration utilities, and backup/restore procedures for self-hosted instances.
Adoption guidance▼
Adopt if
+ You need to deploy a multi-agent business automation workforce but are bound by strict data-privacy regulations (such as the EU AI Act).
+ You want to automate processes across enterprise platforms (like Salesforce, HubSpot, QuickBooks) but require strict outcome validation and human-in-the-loop oversight.
+ You want to use AI agents to dynamically generate and modify workflows using natural language without writing boilerplate orchestration code.
Avoid if
– You are looking for a simple, single-agent personal coding or messaging assistant without enterprise governance, tool sandboxing, or multi-agent structures.
– You prefer a purely visual drag-and-drop workflow tool with zero coding or self-hosting responsibilities.
How it works & getting started▼
How it works
1.The user self-hosts the Atom platform on their own hardware or cloud server.
2.The user configures LLM providers using their own API keys or hooks up a local model provider like Ollama.
3.The user describes a desired business workflow in plain language via a conversational chat interface.
4.Atom automatically authors and deploys the structured workflow, placing default-on human-in-the-loop approval gates before critical steps.
5.Specialized agents execute the workflow steps, routing actions inside an execution sandbox and communicating laterally via Agent Radio.
Getting started
→ Clone the repository to your self-hosted infrastructure.
→ Configure the backend/.env file with your choice of LLM provider keys, an OPENCODE_API_KEY, or set ATOM_LOCAL_ONLY=true and configure OLLAMA_BASE_URL.
→ Launch the system and open http://localhost:3001 in your browser.
→ Retrieve the admin password located in backend/logs/bootstrap_admin_password.txt and log in using the email [email protected].
→ Describe your target workflow in plain language to start building your first automation.
Agent handoff
Use with any agent
JSON API
Alternatives

Nearby repositories worth comparing before adoption.

Compare top options →
Financial-API
HiThink-Tech/Financial-API
59
Fit

同花顺官方 A股金融数据服务，提供股票实时行情、历史行情、财务报表、指数、板块、涨停等数据，适用于 AI Agent、量化研究和应用开发，支持 API、MCP、CLI 和 Python。Official Tonghuashun (HiThink) A-share financial data service providing real-time and historical stock market data, financial statements, indices, sectors and limit-up data for AI agents, quantitative research and application development.

Trust REVIEW · 90
Compare →
holaOS
holaboss-ai/holaOS
71
Fit

holaOS คือ Agentic OS ที่ทำให้ AI Agent สามารถทำงานดิจิทัลได้ทุกอย่างบนคอมพิวเตอร์ของคุณ ไม่ว่าจะเป็นการจัดการไฟล์ ท่องเว็บ หรือรันโปรแกรมต่างๆ ผ่าน Natural Language ด้วย Electron + TypeScript + MCP Protocol ที่ Open Source สำหรับ Developer ที่อยากสร้าง Desktop AI Agent ของตัวเอง

Trust REVIEW · 45
Compare →
TencentDB-Agent-Memory
TencentCloud/TencentDB-Agent-Memory
70
Fit

TencentDB Agent Memory is a team-level memory hub for AI Agents — turning conversations, docs, and code into four reusable memory assets (Chat Memory, Skill, LLM-Wiki, Code-Graph) that are governed, shared, and equipped across agents and frameworks.

Trust REVIEW · 0
Compare →