Version amnesia.
That prompt that worked perfectly three weeks ago? Gone. You rewrote it, forgot the original, and now you're debugging a regression you can't even trace.
Prompt Manager 2.0
The Central Command for AI Architects.
Prompt Manager 2.0 gives engineers a local-first, version-controlled workspace to build, iterate, and deploy prompts with zero API cost and zero vendor lock-in.
One-time payment. No subscription. No expiry.
- Local-first. Data stays on your machine.
- LM Studio & Ollama compatible. $0 per token.
- Lifetime access. One payment.
Version control for the prompts that run your stack.
Your prompt library is a graveyard.
Context collapse.
Slack threads. Notion pages. .txt files in ~/Desktop/misc/. Your best prompts are scattered across tools that don't talk to each other.
API cost creep.
Every iteration loop costs tokens. Testing, tweaking, re-running — it adds up. There's no way to prototype locally when every call hits your bill.
Zero reproducibility.
You got a great output once. You have no idea what prompt configuration produced it. You can't reproduce it. You can't ship it.
Prompt engineering without version control is like shipping code without git.
Prompt Manager 2.0.
Atomic version control. Local-first. Zero API cost.
Prompt Manager 2.0 treats your prompts as what they actually are: load-bearing infrastructure. Every iteration is versioned. Every deployment is tracked. Every rollback is one click.
It runs entirely on your machine. Pair it with LM Studio or Ollama and your entire prompt iteration loop costs $0 per token. No cloud dependency. No rate limits. No vendor extracting rent from your intelligence.
This is Sovereign Intelligence. Your models, your prompts, your rules.
Atomic Versioning
Every prompt change is a discrete, named commit. Diff two versions. Roll back instantly. Ship the exact configuration that worked.
Local-First Architecture
All data stays on your machine. Privacy First. We only use basic Google Analytics to see which prompts are popular and improve the site. No personal data is collected. No sync to external servers. Your prompts don't fund someone else's training set.
Zero Friction Workflow
Open. Edit. Deploy. The interface disappears. The output is all that remains. No unnecessary modals, no onboarding wizards, no "welcome back" emails.
LM Studio & Ollama Integration
Connect any local LLM in under 60 seconds. Run Llama, Mistral, Phi, or any GGUF model. Zero API cost. Infinite iterations.
Built for engineers who think in systems.
Version History & Diff View
Every save creates a versioned snapshot. Compare any two versions side-by-side. See exactly what changed, when, and why.
Prompt Deployment Workflow
Mark a prompt version as "deployed". Track which version is live. Never wonder what's running in production again.
Local LLM Integration
Native LM Studio and Ollama support. Connect your local model endpoint in one line. Full prompt testing loop at $0 per run.
Tag & Context System
Tag prompts by model, use case, project, or status. Filter in milliseconds. Find the exact configuration you need without grep.
Export & Portability
Export any prompt or full library as JSON. Import on a new machine in seconds. Your library moves with you — not with the app.
Offline-First Operation
Zero network dependency for core features. Works on a plane, in a datacenter, behind a firewall. Your workflow never waits for a server.
Privacy First. We only use basic Google Analytics to see which prompts are popular and improve the site. No personal data is collected. No accounts. No expiry dates.
One-time. Lifetime. Yours.
Prompt Manager 2.0 — Lifetime Access
$2.99
One-time payment via Gumroad. No subscription. Ever.
- Lifetime license — one payment, forever
- All future updates included
- Full version history & diff view
- LM Studio & Ollama integration
- Export / import library as JSON
- Offline-first, local-first operation
- Privacy First. We only use basic Google Analytics to see which prompts are popular and improve the site. No personal data is collected.
The cost of a single GPT-4 API call. Paid once. Used forever.
Why not free?
Free tools don't survive. This one will.
Is there a subscription?
No. $2.99 once. That's the transaction.
What if I need a refund?
Gumroad handles all transactions.
Local LLM connected in under 5 minutes.
No API key. No cloud account. No cost per token.
Prompt Manager 2.0 pairs with LM Studio or Ollama to give you a complete, cost-free prompt engineering environment. Here's how to connect in four steps.
Option A — LM Studio
01
Download LM Studio
Go to lmstudio.ai and download the installer for your OS.
Free download. No account needed.
02
Load a Model
Open LM Studio → Search for a model such as Llama 3, Mistral 7B, or Phi-3 → Click Download.
Models run locally. GGUF format recommended for performance.
03
Start the Local Server
In LM Studio, click the Local Server tab → Select your loaded model → Click Start Server.
Default endpoint: http://localhost:1234/v1
LM Studio exposes an OpenAI-compatible API.
04
Connect in Prompt Manager 2.0
Open Prompt Manager 2.0 → Settings → LLM Endpoint → paste your local server URL → Save.
http://localhost:1234/v1
Your prompts now run on local compute. $0 per token.
Option B — Ollama
01
Install Ollama
Visit ollama.com and run the one-line installer for your OS.
curl -fsSL https://ollama.com/install.sh | sh
macOS, Linux, and Windows supported.
02
Pull a Model
Open your terminal and pull any supported model.
ollama pull llama3
Model downloads to local storage. Runs offline after pull.
03
Ollama Runs Automatically
Ollama starts a local server automatically after install. No configuration needed.
Default endpoint: http://localhost:11434
04
Connect in Prompt Manager 2.0
Open Prompt Manager 2.0 → Settings → LLM Endpoint → enter Ollama endpoint → Save.
http://localhost:11434
Same workflow. Zero cost. Full local control.
Your entire prompt development loop — write, version, test, deploy — now runs locally. Forever free to iterate.
Run the local loop.
A working prompt optimizer wired to LM Studio at http://127.0.0.1:1234/v1/chat/completions.
Command Input
Write your prompt here...
Optimized Output
Copy. Deploy. Roll back when needed.
No prompts yet. Create your first deployment.
Connects to LM Studio or Ollama. Zero API cost per run.