null-lens 0.1.0

The missing protocol between user requests and AI actions. Standardized intent parsing for every AI system.

⟁ Null Lens

The missing protocol between user requests and AI actions.
Standardized intent parsing for every AI system.
Converts any input into a deterministic schema — [Motive] [Scope] [Priority].

---

🧩 What it does

Null Lens turns unstructured natural-language input into a fixed three-line intent block:

[Motive] — what the user wants to achieve
[Scope] — where the request applies (context or domain)
[Priority] — what must be done first or resolved

One ambiguous paragraph in → three structured fields out.

---

⚙️ Quick Start

Install

pip install null-lens

Get your API key

You’ll need an API key to use Lens.

→ https://null-core.ai → sign up → get 100 free queries instantly.
No credit card. No approval delay.

export NULL_LENS_API_KEY="your_api_key_here"
# or on Windows PowerShell
setx NULL_LENS_API_KEY "your_api_key_here"

Use

from null_lens import NullLens

lens = NullLens(api_key="your_api_key_here")

result = lens.parse("Summarize Q4 strategy across LATAM markets.")
print(result)

Response

[Motive] Summarize strategic direction for Q4  
[Scope] LATAM markets  
[Priority] Identify key actions for planning cycle

---

🧠 Why it matters

AI systems don’t fail at inference — they fail at interpretation. Lens removes ambiguity before reasoning begins.

Without Lens

• Prompt retries
• Context drift
• RAG scaffolding debt

With Lens

• Stable input schema
• Consistent reasoning
• Deterministic orchestration

---

🔌 API Access

Endpoint

POST https://null-core.ai/api/lens

Headers

Authorization: Bearer <API_KEY>  
Content-Type: application/json

Body

{
  "messages": [
    {
      "role": "user",
      "content": "Summarize the latency impact of our RAG pipeline for 100k qps"
    }
  ]
}

Response

{
  "object": "chat.completion",
  "org_id": "xxxx-xxxx-xxxx",
  "response": "[Motive] Identify latency impact of RAG pipeline\n[Scope] RAG pipeline, 100k QPS scenario\n[Priority] Optimize for performance stability"
}

---

🧩 Parsing Helpers

JavaScript

function parseLensResponse(responseText) {
  const lines = responseText.split(/\r?\n/).map(l => l.trim()).filter(Boolean);
  const find = (prefix) => {
    const line = lines.find(l => l.toLowerCase().startsWith(prefix));
    return line ? line.split(']').slice(1).join(']').trim() : null;
  };
  return {
    motive: find('[motive]'),
    scope: find('[scope]'),
    priority: find('[priority]')
  };
}

Python

def parse_lens_response(text):
    lines = [l.strip() for l in text.splitlines() if l.strip()]
    def find(prefix):
        for l in lines:
            if l.lower().startswith(prefix):
                return l.split(']', 1)[1].strip()
        return None
    return {
        "motive": find("[motive]"),
        "scope": find("[scope]"),
        "priority": find("[priority]")
    }

---

🧩 Example cURL

curl -X POST https://null-core.ai/api/lens \
  -H "Authorization: Bearer <API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{"messages":[{"role":"user","content":"We migrated API to v2 and uptime dropped — logs show auth timeouts. Root cause & mitigation before Friday?"}]}'

---

🧠 Best Practices

• One call per semantic input — send full text in messages[0].content.
• Persist outputs — store Motive / Scope / Priority as fields in your DB.
• Cache identical inputs — identical input → identical output.
• Use Lens before RAG / agents — treat its output as your canonical intent layer.
• Test with messy inputs — long emails, transcripts, logs; the schema holds.

---

🔐 Security & Compliance

• Stateless — no inputs stored or retained.
• You manage your own API keys.
• Rotate keys if compromised.
• Don’t send confidential or PII data unless necessary.
• Provided as-is, without warranties; you own compliance and handling.

---

❓ FAQ

Q: Is Lens stateful? A: No — each call is independent. Inputs are transient and not stored.

Q: Is the output deterministic? A: Yes. Same input → same 3-line output. Determinism by design.

Q: Will the format ever change? A: Never. The [Motive] / [Scope] / [Priority] schema is permanent.

Q: Average latency? A: ~1–5 s per call, depending on region and load.

Q: Enterprise pricing? A: Contact [e-mail] for volume or dedicated environments.

---

🩶 License

MIT License — see LICENSE

---

API-first. Stateless. Deterministic. Every call, every time.