Integrations

Connect FowyldAI to your existing tools and workflows — no code changes required for OpenAI-compatible clients.

OpenAI SDK (Drop-in Replacement)

FowyldAI implements the OpenAI API specification. Point any OpenAI-compatible library to your local instance:

Python

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/v1",
    api_key="not-needed"  # or your FowyldAI API key
)

response = client.chat.completions.create(
    model="fowyld-default",
    messages=[
        {"role": "user", "content": "Summarize this report."}
    ]
)
print(response.choices[0].message.content)

JavaScript / TypeScript

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'http://localhost:8000/v1',
  apiKey: 'not-needed'
});

const response = await client.chat.completions.create({
  model: 'fowyld-default',
  messages: [{ role: 'user', content: 'Explain this error.' }]
});
console.log(response.choices[0].message.content);

LangChain

from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    base_url="http://localhost:8000/v1",
    api_key="not-needed",
    model="fowyld-default"
)

response = llm.invoke("What are the OWASP Top 10?")
print(response.content)

VS Code Extension

The FowyldAI VS Code extension provides inline completions, code review, and chat powered by your local instance.

Install the FowyldAI extension from the VS Code marketplace
Set the engine URL in settings: "fowyldai.engineUrl": "http://localhost:8000"
Use Ctrl+Shift+F to open the FowyldAI chat panel

Fully local The VS Code extension communicates only with your local FowyldAI instance. No data is sent to any external service.

MCP (Model Context Protocol)

FowyldAI exposes a full MCP server for integration with AI agents, coding assistants, and tool-using workflows:

{
  "mcpServers": {
    "fowyldai": {
      "url": "http://localhost:8000/mcp",
      "transport": "streamable-http"
    }
  }
}

Available Cmdlets

Category	Cmdlet	Description
Code	`code_review`	Review code for security, performance, correctness, and style
	`explain_code`	Explain what a code snippet does
	`find_bugs`	Detect bugs and potential issues
	`refactor`	Refactor for readability, performance, DRY, or SOLID
	`generate_code`	Generate code from a natural language description
	`write_tests`	Generate unit tests (pytest, jest, vitest, unittest)
	`generate_docstring`	Generate docstrings for functions and classes
Dev Workflow	`commit_message`	Generate conventional commit messages from diffs
	`iterative_codegen`	Multi-step code generation with iterative refinement
	`execute_python`	Execute Python code in a sandboxed environment
Reasoning	`ask`	General-purpose query with multi-model orchestration
	`plan`	Break down a task into an execution plan
	`reason`	Step-by-step logical reasoning
	`classify`	Classify text into categories
	`verify_reasoning`	Validate a chain of reasoning for logical errors
Text	`summarize`	Summarize documents or conversations
	`compress`	Compress large text for efficient context handling
	`decompose`	Break complex queries into sub-problems
Knowledge	`remember`	Store facts in persistent local memory
	`recall`	Retrieve stored knowledge
	`ingest_document`	Ingest documents into the local knowledge base
	`query_knowledge`	Query the local knowledge base
Workspace	`read_workspace_file`	Read files from the local workspace
	`write_workspace_file`	Write files to the local workspace
	`search_workspace`	Search across workspace files
	`scan_directory`	Scan and list directory contents
System	`health`	Engine health check and status
	`cached_answer`	Check cache before running inference (zero-cost)
	`auto_route`	Automatically route queries to the optimal model
	`estimate_tokens`	Estimate token count for a given input
	`eval_quality`	Evaluate response quality and accuracy

Zero-cost local execution All cmdlets run on local models — no API keys, no cloud calls, no per-token costs. Use cached_answer first to check the response cache before running inference.

CI/CD Pipelines

Use FowyldAI in your build pipeline for automated code review and security scanning:

GitHub Actions

- name: Code review with FowyldAI
  run: |
    curl -X POST http://fowyldai:8000/ask \
      -H "Content-Type: application/json" \
      -d "{
        \"query\": \"Review this diff for security issues\",
        \"context\": \"$(git diff HEAD~1)\"
      }"

Monitoring

Integrate FowyldAI health checks with your monitoring stack:

Tool	Integration
Prometheus	Scrape `/metrics` endpoint (Enterprise edition)
Grafana	Pre-built dashboard templates available
Datadog	Custom check against `/health`
Kubernetes	Liveness/readiness probes on `/health`