dhruvpatel848

dhruvpatel848/mcp-chaos

Sandboxed security testing for tool-connected (MCP) AI agents: attack your own agent in a fake world, get evidence-backed failures and an enforceable guard policy.

4 good first / help-wanted issues · Python · last activity Oct 31, 2026

0 stars 0 forks 0 watchers Python Apache License 2.0
agent-security ai-agents llm-security mcp prompt-injection red-team security
4 Open Issues Need Help Last updated: Oct 31, 2026

Open Issues Need Help

View All on GitHub
dhruvpatel848/mcp-chaos
0

Sandboxed security testing for tool-connected (MCP) AI agents: attack your own agent in a fake world, get evidence-backed failures and an enforceable guard policy.

Python
#agent-security#ai-agents#llm-security#mcp#prompt-injection#red-team#security
dhruvpatel848/mcp-chaos
0

Sandboxed security testing for tool-connected (MCP) AI agents: attack your own agent in a fake world, get evidence-backed failures and an enforceable guard policy.

Python
#agent-security#ai-agents#llm-security#mcp#prompt-injection#red-team#security
good first issue
dhruvpatel848/mcp-chaos
0

Sandboxed security testing for tool-connected (MCP) AI agents: attack your own agent in a fake world, get evidence-backed failures and an enforceable guard policy.

Python
#agent-security#ai-agents#llm-security#mcp#prompt-injection#red-team#security
dhruvpatel848/mcp-chaos
0

Sandboxed security testing for tool-connected (MCP) AI agents: attack your own agent in a fake world, get evidence-backed failures and an enforceable guard policy.

Python
#agent-security#ai-agents#llm-security#mcp#prompt-injection#red-team#security