title	DevOps Incident Suite
emoji	🚨
colorFrom	red
colorTo	blue
sdk	gradio
sdk_version	6.14.0
python_version	3.11
app_file	app.py
pinned	false

Multi-Agent DevOps Incident Analysis Suite

A VS Code-ready GenAI project that uses LangChain + LangGraph to analyze DevOps logs and generate remediation guidance, runbooks, Slack notifications, and JIRA tickets.

What This Project Demonstrates

• Multi-agent architecture
• LangGraph orchestration
• LangChain LLM usage
• DevOps/SRE log analysis
• Automated incident classification
• Remediation recommendation
• Runbook generation
• Optional Slack and JIRA integration
• Optional RAG: Chroma vector store + embeddings for past resolutions (from closed Jira sync script)
• Single LLM call for classification and remediation (fewer round trips than separate classify + plan steps)
• Gradio UI

Agent Flow

End-to-end processing from raw logs to outputs (Presidio runs in the Gradio app before LangGraph). The knowledge base is optional: populate it with scripts/sync_closed_jira_to_kb.py so classify_and_remediate can run similarity search before the model call.

Flow summary:

1. parse_logs — Structure events from raw text.
2. classify_and_remediate — Keyword seed classification → RAG retrieve similar closed issues → one structured LLM response (incident fields + steps + remediation narrative using KB context).
3. generate_runbook — Markdown runbook from steps, optional LLM narrative, and embedded RAG excerpts (issue keys, summaries, snippets) for Slack/Jira and the UI.
4. notify_and_ticket — Slack webhook; Jira ticket for P1/P2 with the full runbook body.

Legacy helpers classifier_agent.classify_incident and remediation_agent.recommend_remediation remain in the repo for reuse but are not wired into the default graph.

flowchart TB
    subgraph input["Input"]
        U["Paste or upload .log / .txt"]
    end

    subgraph gate["Privacy"]
        P["Presidio anonymization<br/>PII / network identifiers"]
    end

    subgraph kb["Knowledge base (optional)"]
        SYNC["scripts/sync_closed_jira_to_kb.py<br/>closed Jira issues → summaries"]
        V["Chroma persistent store<br/>data/chroma"]
        SYNC --> V
    end

    subgraph graph["LangGraph pipeline"]
        A["parse_logs"]
        B["classify_and_remediate<br/>RAG retrieve + single LLM JSON"]
        D["generate_runbook<br/>steps + KB section + narrative"]
        N["notify_and_ticket<br/>Slack webhook · Jira API"]
    end

    subgraph output["UI outputs"]
        S["Incident summary"]
        R["Runbook"]
        I["Slack & Jira status"]
        J["Full agent state JSON"]
    end

    V -.->|"similarity search<br/>optional"| B

    U --> P
    P --> A --> B --> D --> N
    N --> S
    N --> R
    N --> I
    N --> J

Loading

Project Structure

devops-incident-suite/
├── agents/
│   ├── log_agent.py
│   ├── heuristics.py
│   ├── classify_remediate_agent.py
│   ├── classifier_agent.py
│   ├── remediation_agent.py
│   ├── runbook_agent.py
│   └── notification_agent.py
├── graph/
│   └── workflow.py
├── integrations/
│   ├── slack.py
│   └── jira.py
├── knowledge/
│   ├── vector_store.py
│   └── jira_kb_sync.py
├── scripts/
│   └── sync_closed_jira_to_kb.py
├── utils/
│   ├── llm.py
│   ├── embeddings.py
│   └── log_anonymization.py
├── tests/
├── data/sample_logs/
├── app.py
├── state.py
├── requirements.txt
├── .env.example
└── README.md

Setup in VS Code on Windows

1. Open folder in VS Code

Open this repository folder (e.g. devops-incident-suite).

2. Create virtual environment

py -m venv .venv

3. Activate virtual environment

.\.venv\Scripts\Activate.ps1

If PowerShell blocks activation, run:

Set-ExecutionPolicy -Scope CurrentUser RemoteSigned

Then activate again:

.\.venv\Scripts\Activate.ps1

4. Install dependencies

py -m pip install --upgrade pip
py -m pip install -r requirements.txt

5. Configure environment

Copy .env.example to .env.

copy .env.example .env

Add your OpenRouter key:

OPENAI_API_KEY=your_key_here
OPENAI_BASE_URL=https://openrouter.ai/api/v1
LLM_MODEL=openai/gpt-4o-mini

The app also works without an API key using fallback rule-based classification.

6. Run the app

py app.py

Open the local URL shown in the terminal, usually:

http://127.0.0.1:7860

Run in Google Colab

Upload the project folder or zip file, then run:

!pip install -r requirements.txt
!python app.py

For public Colab link, change the last line in app.py from:

demo.launch()

to:

demo.launch(share=True)

Deploy to Hugging Face Spaces

1. Create a new Hugging Face Space
2. Select Gradio as SDK
3. Upload all project files
4. Add secrets in Space settings:
   • OPENAI_API_KEY
   • OPENAI_BASE_URL
   • LLM_MODEL
   • optional Slack/JIRA values
5. Hugging Face will install requirements.txt and run app.py