NikeGunn/agentguard

Add agent detector: Aider about 1 hour ago

enhancement help wanted good first issue

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Add agent detector: Cline about 1 hour ago

enhancement help wanted good first issue

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Add agent detector: Continue about 1 hour ago

enhancement help wanted good first issue

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Add secret-redaction rules for more vendor key formats about 1 hour ago

help wanted good first issue detection-engine

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Grow the prompt-injection corpus with new adversarial payloads about 1 hour ago

help wanted good first issue detection-engine

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Record the README demo GIF(s) about 1 hour ago

documentation help wanted good first issue

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Quickstart doc pass: smooth the first five minutes about 1 hour ago

documentation help wanted good first issue

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Add a fresh-install smoke-test make target about 1 hour ago

help wanted good first issue

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Real loop & cost protection: semantic loop detection + token/dollar budgets about 1 month ago

enhancement help wanted security detection-engine epic

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Tool-poisoning & schema rug-pull defense: pin tool definitions, detect drift, scan descriptions about 1 month ago

enhancement help wanted security detection-engine epic

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Stateful session-aware detection: taint tracking, multi-step attack chains & session traces about 1 month ago

enhancement help wanted security detection-engine epic

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Burp Suite–class interception layer: full capture, history, live interceptor & repeater about 1 month ago

enhancement help wanted security detection-engine epic

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Deep prompt-injection detection: defeat encoding, homoglyph, unicode & multilingual bypasses about 1 month ago

enhancement help wanted security detection-engine epic

NikeGunn/agentguard

3

Zero-trust security sidecar for AI agents - MIT, single Go binary, <5ms p99 overhead, protects Claude Code/Cursor/Codex/Gemini CLI/Windsurf

Go

#ai-agents#claude-code#codex#cursor#golang#mcp#model-context-protocol#prompt-injection#security#sidecar#zero-trust

Open Issues Need Help