Welded to one model
Model A- Tools
- Prompts
- Safety rules
- Memory
Myriaxon Tech/AI harness engineering
Build once.
Run on any AI model.
Myriaxon is the layer between your business and the AI models it uses. We build it, deploy it and look after it, so when a better model arrives it is a setting you change, not a project you start over.
- 6providers built in
- Anycompatible API
- 31built-in tools
- 14point benchmark
Works with the models and tools you already use
- Anthropic
- OpenAI-compatible APIs
- Google Gemini
- Vertex AI
- DeepSeek
- OpenRouter
- Netra
- Local models
- MCP connectors
- GitHub
- Notion
- Linear
- PostgreSQL
- Sentry
- Figma
The problem
A new AI model every few weeks. A rebuild every time.
Most AI tooling is welded to a single vendor. When a faster, cheaper or smarter model appears, teams re-wire their tools, rewrite their prompts and re-check their safety rules. Myriaxon puts one stable harness between your work and the models, so the model can change while everything else stays put.
What is a harness? The model is the engine. The harness is everything around it that makes the engine useful at work.
- ToolsWhat it can use and do
- RulesWhat it may and may not do
- MemoryWhat it remembers
- RecordProof of what it did
With the Myriaxon harness
Model A- Tools
- Prompts
- Safety rules
- Memory
Illustration of the idea. Switching models still deserves a test, which is what the benchmark further down is for.
The idea
Like OpenRouter, but for the harness.
OpenRouter gave developers one API for many models. Myriaxon takes the same idea one level up: one harness that works the same way whichever model or API sits underneath. Same tools. Same rules. Same memory. Choose the model per task, per team or per budget, and change your mind whenever you like.
Your work
- Files and data
- Company systems
- Chat apps
- Your team
Myriaxon harness the same for every model
- Tools
- Rules and approvals
- Memory and skills
- Audit log
- Model adapterSpeaks: Anthropic
Any model or API pick one
It adapts to the model for you
A small local model and a large frontier model need different instructions, limits and tools. Myriaxon detects which one it is talking to and adjusts, so every model performs at its best instead of being forced into one setup.
- Working memory
- 64,000 tokens
- Longest reply
- 8,192 tokens
- Instructions it is given
- 6,000 characters
- Tool output it reads at once
- 8,000 characters
- Tools held back to save space
- 1
- Helper agents
- Off
Typical values for each class of model. Detected automatically, and you can set it by hand.
How it works
Follow one task through the harness.
A request arrives. The harness checks the rules, gets the model ready, lets it work with real tools, pauses for a person when it should, and keeps a record of everything.
Step 1 of 6
Request
A task arrives from a person or a schedule.
Why it matters
Illustration of a typical session.
Seamless integration
Plugs into the way your company already works.
No rip-and-replace. Myriaxon connects to your existing software through open connectors, reaches people in the chat apps they already use, and fits your company's sign-in and security.
Connect your software
Open connectors (called MCP) let the AI read from and act in the tools your team already uses. Eighteen are ready to switch on, and any other can be added.
- GitHub
- Notion
- Linear
- Sentry
- PostgreSQL
- SQLite
- Figma
- Git
- Files
- Web search
- Browser
- and more
Reach people where they are
Staff and customers can message an agent on WhatsApp. Only approved phone numbers get in.
- Twilio
- Meta Cloud API
- Any webhook
Fits your security
Use Google Vertex AI with your company's workforce identity, so work is tied to company accounts and billed to your own cloud project.
- Sign-in
- Company account
- Billing
- Your cloud project
- Access
- Set by your admin
Runs where you need it
From a terminal or a built-in web dashboard. The dashboard needs no outside services, so it also runs in closed networks.
- Terminal
- Web dashboard
- Windows
- macOS
- Linux
Make it yours
Add your own tools, reusable playbooks (called skills), and checks that run before or after any action.
- Plugins
- Skills
- Checks
- Custom commands
Start from a working agent
Ready-made templates for the jobs most teams need first. Each one carries its own instructions and knows when to ask a person.
- Research analyst
- Customer support
- Code reviewer
- Data analyst
- Operations monitor
- Design partner
Where Myriaxon goes further
Built for companies that will outlast any one model.
Harnesses built around a single vendor are fine until that vendor is no longer the best choice. Here is what changes when the harness is independent of the model.
| What you need | A single-vendor harness | Myriaxon |
|---|---|---|
| Choice of AI model | Limited to one vendor's models | Six providers built in, plus any compatible API |
| Changing model | Not possible. There is only one | Switch mid-conversation and keep the conversation |
| Fit to the model | The same setup for every model | Adapts instructions, limits and tools to the model's size |
| Choosing the model | You rely on the vendor's own claims | A built-in benchmark you run on your own tasks |
| Closed networks | Needs the vendor's cloud | Dashboard needs no outside services and pairs with local models |
| When a model is busy | Work stops until you retry | Falls back to a backup model of your choice |
"Single-vendor harness" describes agent tools built around one provider's models. The comparison reflects our own review of publicly documented features and may not match every product or version. Myriaxon also includes what you would expect: tools, approvals, undo, memory, an audit trail and extensions.
Choose with evidence
Find the best model for your business. Don't guess.
Every vendor says their model is the best. Myriaxon includes its own benchmark, called IQB-14, that puts the same tasks through any model and scores it on 14 things that matter at work. When a new model is released, run the same test again and see the difference before you switch anything.
What matters most to you?
Example data to show the idea, not real results. A real scorecard is built from your own tasks and the models you are considering.
The 14 things we measure
- Did it work?
- How did it go about it?
- Was it safe?
- 01Finishes the job
- 02Uses tools correctly
- 03Stays within allowed actions
- 04Reasons in a sensible order
- 05Copes when systems change
- 06Gives consistent answers
- 07Handles rewording
- 08Matches approved answers
- 09Covers every point
- 10Stays grounded in your data
- 11Treats everyone fairly
- 12Resists manipulation
- 13Admits what it doesn't know
- 14Protects secrets
Checked, not assumed. Results are verified against what really happened, such as files and exit codes, not the AI's own account of its work.
Repeatable. Results are saved, so a score can be re-checked later without spending anything more.
Honest by design. A score built on incomplete evidence is labelled provisional, never rounded up.
What we do for you
We build it, deploy it and keep it current.
Myriaxon Tech specialises in deploying and building AI harnesses across a business. Bring us a goal, or an existing setup you want to improve.
Custom Myriaxon build
A harness shaped around your company.
- Your tools and data connected
- Your rules and approvals written in
- Your workflows packaged as reusable playbooks
OpenClaw and Hermes deployment
Run an open-source agent such as OpenClaw or Hermes without the guesswork.
- Installed and secured properly
- Connected to your systems and chat apps
- Access control and hand-over training
Model selection and benchmarking
Know which model fits before you commit.
- Your real tasks, run on the candidate models
- A clear scorecard and recommendation
- A re-test whenever a new model arrives
Integration
Connect AI to the systems your team already lives in.
- Databases, tickets, code and documents
- WhatsApp and other chat apps
- Company sign-in and security review
Support and upgrades
Keep it working as models, and your needs, change.
- Training for your team
- Ongoing support
- Model-change reviews, so upgrades stay simple
Not sure where to start?
Tell us what you want AI to do for your business. We will tell you honestly what is realistic, and what it would take.
Start a conversationHow an engagement works
- 1DiscoverWe learn your work, your risks and your goals.
- 2DesignWe choose models, tools and rules, and agree what success looks like.
- 3BuildWe configure Myriaxon and connect your systems.
- 4DeployWe launch it and train your team.
- 5SupportWe keep it current as models and needs change.
- In your cloud
- On your own servers
- In a closed network, with a local model
Real numbers only.We report what was measured and label what was not.
Safe by default.Risky actions ask first, and your company's rule has the final word.
No lock-in.Your tools, your data and your choice of model stay yours.
Coming soon
Myriaxon Lighthouse
See every AI in your company. Govern all of it.
As companies add more models, agents and bots, it gets hard to say what is running, what it costs and what it is allowed to do. Lighthouse is an AI management system: one place to check, assess, manage and govern every model, harness, agent, bot and swarm in your organisation, whoever built it.
- CheckSee every model, agent and bot in use, and what each one connects to.
- AssessScore quality, risk and cost with the same benchmark.
- ManageChange models, permissions and schedules from one place.
- GovernSet company-wide rules and approvals, with a full audit trail.
Myriaxon already gives each deployment layered rules, approvals and an audit trail. Lighthouse is being built to bring that kind of oversight to every AI in the company. Planned capabilities may change before release.
Contact
Tell us what you want AI to do for your business.
Describe the goal, the team, and any models or tools you already use. We will reply by email to arrange a call. Existing customers can choose Support and include their company name.
- You write to usA few lines is enough.
- We replyBy email, to set up a short call.
- We scope itA pilot or a full build, with clear deliverables.