Skip to content

PraisonAI

Untested

AI Agents Framework with Self Reflection and MCP support

by MervinPraison 5,675 MIT Python Intermediate stdio

Install

Claude Code

claude mcp add praisonai -- uvx PraisonAI

Safety Report

60% Untested
Scanned yesterday
3 passed 2 warnings
Security
Authentication
details

Supports auth_token parameter for WebSocket MCP connections. API keys passed via environment variables. No built-in auth for its own MCP server exposure.

Info
CORS Policy
details

HTTP transport supported; CORS configuration depends on deployment setup and is not documented by default.

Info
Rate Limiting
details

No documented rate limiting for self-hosted deployments.

Warning
Known CVEs
details

No known CVEs at time of review.

Info
Code Quality
Dependency Audit
details

Python package published on PyPI. Dependency audit not publicly documented.

Info
Dangerous Patterns
details

Dangerous patterns detected: subprocess_shell

Warning
License
details

MIT license.

Pass
Code Source
details

Open source on GitHub with active development.

Pass
Community
Maintenance Status
details

Actively maintained with frequent commits.

Pass

Learn about our security methodology →

Compatibility

Claude CodeCan be used as an MCP client framework; connect to MCP servers via stdio, HTTP, or WebSocket.
Not Tested
CursorWorks as a framework for building agents that consume MCP tools.
Not Tested
VS CodeWorks as a Python-based agent framework.
Not Tested
WindsurfNo documented direct integration.
Partial
Claude DesktopCan expose an MCP server endpoint for Claude Desktop to connect to.
Not Tested

Tools Provided (4)

run_agent

Executes a PraisonAI agent with specified tools and task instructions.

create_agent

Creates a new AI agent with a defined role, goal, and toolset.

list_agents

Lists all configured agents in the PraisonAI instance.

run_task

Runs a specific task using a configured agent.

Frequently Asked Questions

What is PraisonAI?

AI Agents Framework with Self Reflection and MCP support

What tools does PraisonAI provide?

PraisonAI provides 4 tools including run_agent, create_agent, list_agents, run_task. Each tool can be called by AI agents to perform specific operations.

Is PraisonAI safe to use?

PraisonAI has not yet been security scanned. We recommend reviewing the source code and safety report before use.

What are alternatives to PraisonAI?

Similar MCP servers include FastMCP, Oh My Posh Validator, Claude Flow. Each serves a similar purpose but may differ in features, language, and compatibility.

Similar MCP Servers