Abidemialade

Abidemialade/mylonite

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

1 stars 0 forks 1 watchers Python Apache License 2.0
agent-security ai-security llm-security mcp owasp prompt-injection pytest python red-teaming security-testing
View on GitHub
6 Open Issues Need Help Last updated: Aug 29, 2026

Open Issues Need Help

View All on GitHub
Attack module: multi-turn permission drift into an unconfirmed send (W4) about 2 hours ago
enhancement help wanted good first issue
Abidemialade/mylonite
1

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

Python
#agent-security#ai-security#llm-security#mcp#owasp#prompt-injection#pytest#python#red-teaming#security-testing
Attack module: exfiltration via markdown image/link in tool output (W3) about 2 hours ago
enhancement help wanted good first issue
Abidemialade/mylonite
1

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

Python
#agent-security#ai-security#llm-security#mcp#owasp#prompt-injection#pytest#python#red-teaming#security-testing
Attack module: egress via redirect chain (W3) about 2 hours ago
enhancement help wanted good first issue
Abidemialade/mylonite
1

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

Python
#agent-security#ai-security#llm-security#mcp#owasp#prompt-injection#pytest#python#red-teaming#security-testing
Attack module: persistent write-then-read poisoning (W2) about 2 hours ago
enhancement help wanted good first issue
Abidemialade/mylonite
1

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

Python
#agent-security#ai-security#llm-security#mcp#owasp#prompt-injection#pytest#python#red-teaming#security-testing
Attack module: tool-description rug pull after first approval (W1) about 2 hours ago
enhancement help wanted good first issue
Abidemialade/mylonite
1

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

Python
#agent-security#ai-security#llm-security#mcp#owasp#prompt-injection#pytest#python#red-teaming#security-testing
Attack module: ASCII smuggling via invisible Unicode tag characters (W1) about 2 hours ago
enhancement help wanted good first issue
Abidemialade/mylonite
1

Open-source AI-layer security testing — probes your AI agent for weaknesses and writes a validated regression test for each one it finds, gating CI.

Python
#agent-security#ai-security#llm-security#mcp#owasp#prompt-injection#pytest#python#red-teaming#security-testing