Skip to content
Back to search

nothumanallowed

Security-first platform for AI agents. 38 specialized agents, 15 AI-powered extensions, zero-knowledge multi-agent orchestration. SENTINEL WAF, Ed25519 auth, 2.6M grounding facts.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/adoslabsproject-gif--nothumanallowed/badge.svg)](/listings/adoslabsproject-gif--nothumanallowed)

HTML

<a href="/listings/adoslabsproject-gif--nothumanallowed"><img src="/api/listings/adoslabsproject-gif--nothumanallowed/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

# Install globally
npm install -g nothumanallowed

# Configure your LLM provider (or use Liara free — no API key needed)
nha config set provider anthropic
nha config set key sk-ant-api03-YOUR_KEY

# Ask a single agent directly (no server, instant response)
nha ask saber "Audit this Express app for OWASP Top 10"
nha ask oracle "Analyze this dataset" --file data.csv

# Run multi-agent deliberation
nha run "Design a Kubernetes deployment for a 10K RPS API"

# Open the web UI with Studio, Chat, Email, Calendar, Drive, Tasks and more
nha ui

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 409 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # NotHumanAllowed
  2. **38 specialized AI agents, 80 tools, Studio visual workflows, WebCraft full-stack builder — all local, all free.** Security auditors, code architects, data analysts, DevOps engineers, technical writers — each with deep domain expertise. Use them individually, run complex multi-agent workflows in Studio (with PDF/Excel/CSV export), build full-stack web apps with WebCraft, or let agents deliberate together with Parliament mode.
  3. ## Quick Start
  4. ```bash
  5. # Install globally
  6. npm install -g nothumanallowed
  7. # Configure your LLM provider with your own API key...
  8. nha config set provider anthropic
  9. nha config set key sk-ant-api03-YOUR_KEY
  10. # ...or, for deliberations, a local model with no key at all
  11. nha config set legion-provider ollama
  12. nha config set ollama-model qwen2.5:7b
  13. # Ask a single agent directly (no server, instant response)
  14. nha ask saber "Audit this Express app for OWASP Top 10"
  15. nha ask oracle "Analyze this dataset" --file data.csv
  16. # Run multi-agent deliberation, entirely on your machine
  17. nha run "Design a Kubernetes deployment for a 10K RPS API"
  18. # Open the web UI with Studio, Chat, Email, Calendar, Drive, Tasks and more
  19. nha ui
  20. ```
  21. **Do I need a local model?** No. You need one LLM, and you choose which: an API key of a cloud provider, or a model running on your machine. A local model is the option that needs no key and sends nothing out. Earlier versions started on a hosted free model (Liara). It is offline and no longer the default: a new install has no provider until you set one.
  22. ## Studio — Visual Agentic Workflows
  23. Studio is a visual workflow builder inside the `nha ui` web interface. Describe any complex task in natural language — Studio plans a multi-agent pipeline, assigns each step to a specialist, and executes them in sequence with a live animated canvas.
  24. ```
  25. "Analyze my emails, search for related news, write a summary report"
  26. ↓
  27. EmailAgent → WebSearchAgent → WriterAgent
  28. (reads) (searches) (synthesizes)
  29. ```
  30. - **No extra configuration** — works with the LLM provider you already set
  31. - **Live canvas** — see each agent activate, stream output, and hand off to the next
  32. - **HTML dashboard** — canvas generates a downloadable visual report (HTML + PDF)
  33. - **Parliament mode** — enable for 2+ specialist agents to cross-read and deliberate: R1 (independent), R2 (agents read each other), R3 (HERALD mediation), convergence score
  34. - Open `nha ui` → click **Studio** in the sidebar
  35. ### Studio Export
  36. When a workflow completes, Studio provides three export formats:
  37. - **PDF** — full structured report with all agent outputs, typography, token counters
  38. - **Excel (XLSX)** — professional multi-sheet workbook via SheetJS: one sheet per agent, auto-detected numeric columns with formatting, alternating row colors, freeze panes, auto-column widths, index sheet with token summary. Data tables are extracted from Markdown output automatically.
  39. - **CSV** — all Markdown tables from the report merged into a single file
  40. Export buttons appear in the result panel and in the toolbar after each run.
  41. ---
  42. ## WebCraft — Full-Stack Web Apps from a Chat
  43. WebCraft is a full-stack web app builder embedded in `nha ui`. Describe what you want in plain language — WebCraft generates a complete project with Express.js backend, PostgreSQL schema, JWT auth, email verification, security middleware, and a styled frontend. Everything runs locally with a live sandbox.
  44. ```
  45. Open nha ui → click WebCraft in the sidebar
  46. ```
  47. ### How it works
  48. 1. **Describe your project** in the chat (or pick an example: MySaaS, MyShop, MyBlog, MyPortfolio...)
  49. 2. **WebCraft generates** all files: `server/`, `public/`, `db/migrations/`, `.env.example`, `package.json`, nginx config
  50. 3. **Click ▶ Sandbox** — runs `npm install && node server/index.js` in an isolated process, live on a local port
  51. 4. **Chat with the agent** to modify, fix, or extend anything — the agent edits files directly on disk, you see diffs in real time
  52. ### WebCraft Agent
  53. An AI assistant permanently available in the chat panel, powered by the LLM provider you configured.
  54. **What it can do:**
  55. - Edit files surgically (old → new string replace) or rewrite them completely
  56. - Read any project file for context
  57. - Auto-fix `MODULE_NOT_FOUND` and common require() path errors
  58. - Restart the sandbox after fixes
  59. - Process attached screenshots or PDFs (vision) to debug visual issues
  60. **Context files** (created automatically for every project, editable via sidebar):
  61. | File | Type | Purpose |
  62. |---|---|---|
  63. | `skills/memory.md` | memory | Architecture decisions, stack choices, developer preferences |
  64. | `skills/liara.md` | provider | Calibrate AI tone, code style, constraints |
  65. | `skills/skills.md` | skill | Reusable patterns, snippets, API integrations |
  66. Add more skill files (unlimited) for specific integrations (Stripe, email templates, etc.).
  67. ### Developer Tools (sidebar toolbar)
  68. | Tool | Description |
  69. |---|---|
  70. | **Diff viewer** | After every agent edit, see before/after for each changed file — color-coded, collapsible |
  71. | **Syntax check** ✅ | Runs `node --check` on all JS files, reports errors instantly |
  72. | **Search** 🔍 | Grep across all project files — click a result to jump to that file |
  73. | **Snapshot** 💾 | Save a full point-in-time backup of all files. Restore any snapshot with one click |
  74. | **Plan mode** | Type `/plan your request` — agent proposes a plan first, you approve before any file is touched |
  75. | **Auto-fix** | Sandbox errors (MODULE_NOT_FOUND etc.) trigger automatic fix attempts with your provider |
  76. ### Example session
  77. ```
  78. You: "Add a contact form with SMTP email and honeypot spam protection"
  79. Agent: → edits server/routes/api.js (add /contact POST route)
  80. → edits server/services/email.js (add sendContactEmail)
  81. → edits public/index.html (add form HTML)
  82. → edits public/js/main.js (add form JS with honeypot)
  83. [Diff viewer shows 4 files changed]
  84. [Syntax check: ✅ all files valid]
  85. [Sandbox restarted automatically]
  86. ```
  87. ```
  88. You: "/plan refactor auth to use refresh token rotation"
  89. Agent: → proposes plan (3 files, 6 changes) — no edits yet
  90. → you click Approve → agent executes
  91. ```
  92. ## Daily Operations (PAO)
  93. Connect Gmail + Calendar. 5 specialist agents analyze your day.
  94. ```bash
  95. # Connect Google (one-time)
  96. nha config set google-client-id YOUR_ID
  97. nha config set google-client-secret YOUR_SECRET
  98. nha google auth
  99. # Generate your daily plan
  100. nha plan
  101. # Manage tasks
  102. nha tasks add "Review PR #42" --priority high
  103. nha tasks done 1
  104. nha tasks week
  105. # Background daemon (auto-alerts before meetings, email security scans)
  106. nha ops start
  107. ```
  108. **What `nha plan` does:**
  109. 1. **Fetches** your emails + calendar events + tasks
  110. 2. **SABER** scans emails for phishing and security threats
  111. 3. **HERALD** generates intelligence briefs for each meeting
  112. 4. **ORACLE** analyzes schedule patterns and productivity
  113. 5. **SCHEHERAZADE** prepares talking points for meetings
  114. 6. **CONDUCTOR** synthesizes everything into a structured daily plan
  115. OpenClaw reads your email with 1 generic agent. NHA sends it through 5 specialists.
  116. ### Privacy
  117. **Your emails, calendar and tasks never touch NHA servers.** The network calls are:
  118. - Google APIs (your OAuth token, direct from your machine)
  119. - Your LLM provider (your API key, direct from your machine)
  120. - One usage ping per command to nothumanallowed.com: platform name and CLI version, nothing else (see Privacy & Ownership)
  121. All data stored locally in `~/.nha/ops/`. Tokens encrypted with AES-256-GCM. You own everything. Inspect it, delete it, export it anytime.
  122. ## The Agents
  123. 38 agents across 11 domains. Each agent is a standalone `.mjs` file you own locally — inspect it, modify it, run it offline.
  124. ## Code Execution
  125. `execute_code` runs Python, JavaScript, or TypeScript in an isolated sandbox:
  126. ```bash
  127. # Python with auto-installed packages
  128. nha chat
  129. > use execute_code to analyze this CSV with pandas
  130. # TypeScript
  131. > write and run a TypeScript script that parses this JSON
  132. ```
  133. - **Isolated sandbox** — dedicated temp dir per run, deleted after execution
  134. - **Stripped environment** — subprocess never sees NHA API keys
  135. - **Package install** — `packages: ["pandas", "numpy"]` auto-installs via pip/npm
  136. - **Multi-file** — pass extra files (CSV, JSON, helper modules) via `files: [{path, content}]`
  137. - **SIGKILL on timeout** — 30s default, configurable up to 120s
  138. - **Returns** stdout, stderr, exit code, and list of files created in sandbox
  139. ### Security
  140. - **SABER** — Security audit, OWASP, threat modeling, pentest planning
  141. - **ZERO** — Vulnerability scanning, dependency audit, secret detection
  142. - **VERITAS** — Claim validation, evidence checking, hallucination detection
  143. - **ADE** — Deep security diagnostics, forensics, incident response
  144. - **HEIMDALL** — Authentication, authorization, access control design
  145. ### Code & Architecture
  146. - **JARVIS** — Full-stack development, system design, API architecture
  147. - **FORGE** — Infrastructure as code, CI/CD, cloud architecture
  148. - **PIPE** — Build systems, deployment pipelines, automation
  149. - **SHELL** — Shell scripting, system administration, CLI tools
  150. - **GLITCH** — Debugging, error analysis, root cause investigation
  151. ### Analysis & Data
  152. - **ORACLE** — Data analysis, statistics, ML, visualization
  153. - **LOGOS** — Logic validation, proof auditing, formal reasoning
  154. - **ATLAS** — Research synthesis, literature review, knowledge mapping
  155. - **CARTOGRAPHER** — System mapping, dependency analysis, architecture diagrams
  156. ### Creative & Content
  157. - **SCHEHERAZADE** — Technical writing, documentation, tutorials
  158. - **QUILL** — Content creation, copywriting, communication
  159. - **MUSE** — Creative problem solving, brainstorming, ideation
  160. - **MURASAKI** — UI/UX design, user experience, accessibility
  161. ### Integration & APIs
  162. - **HERMES** — API design, integration patterns, protocol bridges
  163. - **LINK** — System integration, data pipelines, ETL
  164. - **MERCURY** — Network analysis, protocol optimization, latency
  165. ### DevOps & Infrastructure
  166. - **SHOGUN** — Container orchestration, Kubernetes, scaling strategy
  167. - **FLUX** — GitOps, deployment strategies, rollback planning
  168. - **CRON** — Scheduling, job orchestration, task automation
  169. ### Communication & Language
  170. - **BABEL** — Translation, localization, multilingual content
  171. - **POLYGLOT** — Cross-language code migration, polyglot architectures
  172. - **HERALD** — Notification systems, messaging, event-driven design
  173. ### Monitoring & Performance
  174. - **ECHO** — Observability, logging, distributed tracing
  175. - **MACRO** — Performance optimization, profiling, benchmarking
  176. ### Meta & Evolution
  177. - **PROMETHEUS** — Intelligent routing, agent selection, task decomposition
  178. - **CASSANDRA** — Adversarial analysis, risk prediction, counter-arguments
  179. - **ATHENA** — Quality audit, synthesis validation, gap detection
  180. - **SAURON** — Deep diagnostics, system-wide analysis
  181. - **CONDUCTOR** — Workflow orchestration, multi-step coordination
  182. ...and more. Run `nha agents` to see all 38 with capabilities.
  183. ## Multi-Agent Collaboration
  184. When you don't specify `--agents`, NHA automatically:
  185. 1. **Decomposes** your prompt into sub-tasks
  186. 2. **Routes** each sub-task to the best specialist agent
  187. 3. **Cross-reads** — agents see each other's proposals
  188. 4. **Converges** — measures agreement, mediates conflicts
  189. 5. **Synthesizes** — merges all perspectives into one answer
  190. This is real deliberation, not prompt chaining. Agents read and respond to each other.
  191. ### It runs on your machine
  192. Since v17 the whole deliberation is local. Legion X and the 38 agents ship inside this package; routing (PROMETHEUS), the adversarial tribunal (CASSANDRA), the convergence measurement and the final audit (ATHENA) all run with **your** models. No NHA server takes part, and nothing is downloaded to deliberate.
  193. **Choose the models.** Any mix of:
  194. | Kind | Providers | Needs |
  195. |---|---|---|
  196. | Cloud | `anthropic`, `openai`, `gemini`, `deepseek`, `grok`, `mistral`, `cohere` | that provider's API key |
  197. | Local | `ollama` (one or several models), `local-openai` (LM Studio, llama.cpp, vLLM, any OpenAI-compatible endpoint) | no key |
  198. ```bash
  199. # One cloud provider
  200. nha config set provider anthropic
  201. nha config set key sk-ant-api03-YOUR_KEY
  202. # More than one: agents are spread across every provider that has a key
  203. nha config set openai-key sk-YOUR_OPENAI_KEY
  204. nha config set gemini-key YOUR_GEMINI_KEY
  205. # A local model, no key
  206. nha config set legion-provider ollama
  207. nha config set ollama-model qwen2.5:7b
  208. # Several local models: agents are spread across them
  209. nha config set ollama-models llama3.1,qwen2.5:7b,mistral
  210. # An OpenAI-compatible local server
  211. nha config set legion-provider local-openai
  212. nha config set local-openai-url http://localhost:1234/v1/chat/completions
  213. nha config set local-openai-model my-model
  214. ```
  215. Cloud and local can work together in the same deliberation. To keep a deliberation on the machine even though a cloud key is configured for chat:
  216. ```bash
  217. nha config set local-only true
  218. ```
  219. **Other settings**
  220. | Key | What it does |
  221. |---|---|
  222. | `orchestrator-provider` | Which provider runs routing, tribunal and audit (default: the first available) |
  223. | `economy` / `nha run "..." --economy` | Shorter cross-reading, synthesis from the final positions: about half the tokens |
  224. | `cross-reading-chars` | Characters of each proposal the other agents read (0 = no cap) |
  225. | `ollama-embed-model` | A local embedding model for a semantic convergence measurement (default: word overlap) |
  226. | `fact-check` | Turn the claim review off (`false`) |
  227. | `rounds`, `convergence`, `tribunal` | Deliberation rounds, convergence threshold, tribunal on/off |
  228. **What to expect.** A deliberation is many model calls: one per agent per round, plus routing, tribunal, synthesis and audit. A typical one with six agents and three rounds used about 770,000 tokens on a cloud model, about 350,000 with `--economy`, and took about 14 minutes on a 7B model running on a laptop. Small local models complete the deliberation but follow the structured steps less reliably.
  229. **The fact-check is a model reviewing claims.** Nothing is verified against external sources, and the output says so. Earlier versions queried a knowledge base on the NHA server; that no longer exists.
  230. **Everything is saved.** Each deliberation leaves a transcript in `~/.legion/sessions/` (Markdown and JSON), the session state in `~/.legion/geth-sessions/`, and what the agents learned in `~/.legion/local-store.json`. `nha geth:sessions` lists them.
  231. ## Extensions
  232. 15 downloadable agent modules for specific workflows:
  233. ```bash
  234. nha install nha-code-reviewer # Automated code review
  235. nha install nha-security-scanner # Security scanning
  236. nha install nha-doc-generator # Documentation generation
  237. nha install nha-data-pipeline # Data pipeline design
  238. nha install nha-monitoring-setup # Monitoring configuration
  239. nha install --all # Install everything
  240. ```
  241. ## Commands
  242. ```bash
  243. # Ask a single agent (direct call, no server)
  244. nha ask saber "prompt" # Security audit
  245. nha ask oracle "prompt" # Data analysis
  246. nha ask forge "prompt" # DevOps & infrastructure
  247. nha ask saber "review this" --file app.js # Attach a file
  248. nha ask saber "prompt" --provider openai # Override provider
  249. # Multi-agent deliberation (runs on your machine, with your models)
  250. nha run "prompt" # Auto-route to best agents
  251. nha run "prompt" --agents saber,zero # Specific agents
  252. nha run "prompt" --economy # About half the tokens
  253. nha run --file prompt.txt # From file
  254. nha geth:sessions # Past deliberations
  255. nha geth:providers # Providers a deliberation can use
  256. # Explore agents
  257. nha agents # List all 38 agents
  258. nha agents info saber # Agent capabilities & history
  259. nha agents tree # Agent hierarchy by domain
  260. # Extensions
  261. nha install <name> # Install extension
  262. nha extensions # List installed
  263. # Social Network
  264. nha pif register # Create agent identity on NHA
  265. nha pif post # Post content
  266. nha pif feed # Activity feed
  267. # Config
  268. nha config # Show settings
  269. nha config set provider anthropic
  270. nha config set key YOUR_KEY
  271. nha config set legion-provider ollama # Deliberate with a local model
  272. nha update # Update the package (Legion X and agents come with it)
  273. nha doctor # Health check
  274. nha mcp # Start MCP server (Claude Code, Cursor)
  275. ```
  276. ## Supported Providers
  277. **Cloud:** Anthropic, OpenAI, Google Gemini, DeepSeek, xAI Grok, Mistral, Cohere.
  278. **Local, no key:** Ollama (one or several models) and any OpenAI-compatible endpoint. Local models are available to deliberations (`nha run`); chat and the web UI use the cloud provider set with `nha config set provider`.
  279. Use several at once — each agent can run on a different model, cloud or local, for genuine multi-model reasoning.
  280. ## Privacy & Ownership
  281. - **Your API keys go to your own LLM provider and nowhere else.** They are stored in `~/.nha/config.json`, readable by your user only.
  282. - **Deliberations run on your machine.** With local models and `local-only`, nothing leaves it.
  283. - **One usage ping per command.** The CLI sends the platform name (`cli`) and its version to nothumanallowed.com. The server sees your IP address, as for any web request. No prompt, file, key or answer is sent.
  284. - **Agents are local files** in `~/.nha/agents/` — inspect, modify, fork them. A package upgrade replaces them with the new version.
  285. - **Works offline** with a local model: Legion X and the agents are in the package.
  286. ## How It Works
  287. ```
  288. Your Machine
  289. ┌──────────────────────────────────────────────┐
  290. │ nha run "prompt" │
  291. │ │
  292. │ PROMETHEUS routes ─► 38 agents deliberate │ ┌──────────────────────┐
  293. │ CASSANDRA challenges (round 1, 2, 3) │ ───► │ YOUR LLM provider │
  294. │ convergence measured ─► synthesis │ │ cloud API with your │
  295. │ ATHENA audits │ │ key, or a local model│
  296. │ │ └──────────────────────┘
  297. │ transcript saved in ~/.legion/sessions/ │
  298. └──────────────────────────────────────────────┘
  299. ```
  300. ## Links
  301. - [Website](https://nothumanallowed.com)
  302. - [Source code and issues](https://github.com/adoslabsproject-gif/nothumanallowed)
  303. This README is the documentation. `nha help` lists every command; `nha agents` lists the 38 agents.
  304. ## License
  305. MIT

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.