Skip to main content

Nyra 🧠

"Nyra: Because AI should reason, not just respond."

Made by Somya Bhalani, Founder, Ananta Labs (www.anantalabs.app)

Nyra is a deterministic, highly structured Multi-Agent Cognitive Architecture built purely in Python with zero external dependencies. It is not just another chatbot wrapper; it is a profound execution engine that models the reasoning, reflection, and delegation required for autonomous AI.

Architecture Highlights

Nyra breaks down autonomous task execution into distinct, strictly typed cognitive modules:

  • Pre-Frontal Cortex (Planner): Deconstructs complex goals into a DAG of sequential tasks.
  • Reasoner: Selects the absolute best exact action (Tool execution or Delegation).
  • Reflection Engine (Critic): Analyzes the results of actions. If a tool fails, it forces the Reasoner to try again or delegate.
  • Multi-Agent Orchestrator: Dynamically routes sub-tasks to specialized Agent profiles.
  • 3-Tiered Memory: Fully isolates WorkingMemory, EpisodicMemory, and SemanticMemory.

Installation

Because Nyra is purely built on standard Python libraries, installation is incredibly lightweight:

pip install nyra

Quick Start (Bring Your Own API)

Nyra is 100% LLM agnostic. You plug in your own API connection (OpenAI, Gemini, NVIDIA, Claude, etc.) by subclassing the LLM object.

from nyra import Agent
from nyra.llm import LLM
import urllib.request
import json

# 1. Write your custom Provider
class MyCustomLLM(LLM):
    def __init__(self, api_key: str):
        super().__init__(provider="custom")
        self.api_key = api_key

    def generate(self, prompt: str) -> str:
        # Put your API connection logic here!
        return "I am thinking..."

# 2. Boot up the brain
agent = Agent()
agent.llm = MyCustomLLM(api_key="your-key")

# 3. Add tools with Type Hinting (Nyra automatically extracts signatures!)
def calculate_gravity(mass: float) -> float:
    """Calculates gravitational force."""
    return mass * 9.8

agent.add_tool("gravity_calc", calculate_gravity)

# 4. Run the deterministic loop
agent.run("Calculate the gravity for a 50kg object.")

Contributing

Built with passion by Ananta Labs. We believe the future of AI is deterministic, modular, and open. PRs are welcome!

Metadata

Release files for nyra 0.1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for nyra 0.1.0
File Size Uploaded Provenance Actions
nyra-0.1.0.tar.gz 11.2 kB Without provenance Details for nyra-0.1.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for nyra 0.1.0
File Interpreter ABI Platform Actions
nyra-0.1.0-py3-none-any.whl Python 3 none any Details for nyra-0.1.0-py3-none-any.whl

Total release size: 24.9 kB

Release files / nyra-0.1.0.tar.gz

Download URL nyra-0.1.0.tar.gz
Size 11.2 kB
Tags Source
SHA-256 checksum
How to use checksums
b89646fab8ad78a919fdb10fcfc61ae827e22479f3df830f53faf3b88d35125e
BLAKE2b-256 checksum
How to use checksums
706c4535be6cb0241b95728786da0c966910ee3d0bc2194223579b3ff5708942
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.10.8

Release files / nyra-0.1.0-py3-none-any.whl

Download URL nyra-0.1.0-py3-none-any.whl
Size 13.7 kB
Tags Python 3
SHA-256 checksum
How to use checksums
3c21f184d2c767aad3e5b94b78537b9169f7d2d99175e7c368161b995e876e81
BLAKE2b-256 checksum
How to use checksums
3ae4320ae1e48baae0b4f418e8dbaf7f04c7721256a2ab6171f5628dbfb10ed9
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.10.8

Security for nyra 0.1.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page