Skip to main content

Enterprise IT Helpdesk Assistant CLI with RAG, memory, and escalation workflows.

Project description

AegisDesk

Enterprise Autonomous IT Intelligence. Not just a chatbot wrapper.
A deterministic, zero-token semantic routing swarm designed for zero-trust enterprise IT.

Most systems use slow, monolithic LLM calls that hallucinate and burn API tokens. AegisDesk uses a persistent SQLite Graph Memory, sub-5ms local embeddings, and isolated sub-agents to resolve IT tickets autonomously.

Python 3.12+ LangGraph SQLite


The Problem with RAG in the Enterprise

Your AI forgets context when the window closes, hallucinates network commands, and burns through thousands of tokens just trying to figure out if it should query a database or ping a server. Legacy systems rely on "prompt engineering" a massive, monolithic agent.

The AegisDesk Solution

AegisDesk abandons the "monolithic prompt" anti-pattern. Instead, incoming queries are routed through a hyper-optimized deterministic router directly to specialized worker agents.

  1. Zero-Token Semantic Router: Uses local FastEmbed (BAAI/bge-small-en-v1.5) to embed and route intents in < 5 milliseconds without hitting an LLM.
  2. Multi-Agent Swarm:
    • Network Agent: Executes OS-level diagnostics (Ping, Port Scans) with strict Regex RCE sanitization.
    • Cloud Agent: Interfaces directly with Azure/AWS, Atlassian, and Okta via secured APIs.
    • Web Intelligence: Autonomously navigates and scrapes internal wikis. Protected against SSRF via DNS IP resolution filters.
  3. ACID-Compliant Graph Memory: Tracks entities and relationships persistently in SQLite.

📊 Enterprise Benchmarks (50-Query Stress Test)

We subjected AegisDesk's core semantic routing engine to a rigorous 50-query IT Helpdesk simulation.

  • Avg Latency: 4.47 ms per query (Zero-Bloat ONNX Inferencing)
  • Token Cost: $0.00 (100% Local Inference for routing)
  • Direct Match Rate: 70% exact sub-agent routing. The remaining 30% dynamically fall back to the secure RAG evaluator, guaranteeing a 100% resolution pipeline without dropping queries.

🛠️ Quick Start

Installation

You can install AegisDesk directly from PyPI:

pip install aegisdesk

Developer Installation (From Source)

git clone https://github.com/sitanshukr08/Aegisdesk.git
cd Aegisdesk
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install -e .

Initialization

# Initialize data structures, logs, and environments
aegisdesk init

# Ingest HR / IT Documentation into the ChromaDB Vector Store
aegisdesk ingest ./docs/vpn_troubleshooting.pdf

CLI Execution

AegisDesk features a beautiful, Rich-powered interactive CLI for headless server deployments.

aegisdesk ask "Can you ping the corporate gateway and check if my Okta token expired?"

🛡️ Zero-Trust Security Protocols

AegisDesk is hardened against Red Team exploits:

  • RCE Prevention: shell=True is explicitly disabled. All OS inputs are stripped of shell metacharacters (&, |, ;, $, <).
  • SSRF Mitigation: All web scraper requests undergo pre-flight DNS resolution. Any attempt to scrape private, loopback, or link-local subnets aborts instantly.
  • Denial of Wallet: The LangGraph Supervisor dynamically counts recursive agent tool_calls and explicitly halts infinite loops.
=============================== tests coverage ================================
Name                                      Stmts   Miss  Cover
-------------------------------------------------------------
app\rag\graph.py                            120     62    48%
app\rag\pipeline.py                          83     40    52%
src\aegisdesk\core\llm_factory.py            29      4    86%
src\aegisdesk\core\web_tools.py              70     15    79%
-------------------------------------------------------------
TOTAL                                      1218    729    40%

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

aegisdesk-0.1.5.tar.gz (21.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

aegisdesk-0.1.5-py3-none-any.whl (18.8 kB view details)

Uploaded Python 3

File details

Details for the file aegisdesk-0.1.5.tar.gz.

File metadata

  • Download URL: aegisdesk-0.1.5.tar.gz
  • Upload date:
  • Size: 21.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.10

File hashes

Hashes for aegisdesk-0.1.5.tar.gz
Algorithm Hash digest
SHA256 2471476ed9d1306cd9e6be00e9c85543187623630c4d42913a24339ba53251ce
MD5 f3418ccc0c039485a35837f1e35a0a76
BLAKE2b-256 df0f94aef2a821735b4a6fc99f909210a911411165af8f0fda611f86078ddf01

See more details on using hashes here.

File details

Details for the file aegisdesk-0.1.5-py3-none-any.whl.

File metadata

  • Download URL: aegisdesk-0.1.5-py3-none-any.whl
  • Upload date:
  • Size: 18.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.10

File hashes

Hashes for aegisdesk-0.1.5-py3-none-any.whl
Algorithm Hash digest
SHA256 e67c375e382301984278cf4401331553864154120a99e126e853db9dcb11309d
MD5 18c5c9e54b06147e76cd3a426ccf4341
BLAKE2b-256 369aad25f65c342da76a6a05af3c04e3c476a5293fcd47ebae53473817cf1927

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page