Open Issues Need Help
View All on GitHub Translate README to Simplified Chinese (README.zh-CN.md) about 1 month ago
good first issue
empire-mind/agent-guardian empire-mind/agent-guardian A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Open project →
0
A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Python
#ai-agents#cli#devtools#python#security
False-positive audit table for the 12 MEDIUM patterns about 1 month ago
help wanted
empire-mind/agent-guardian empire-mind/agent-guardian A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Open project →
0
A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Python
#ai-agents#cli#devtools#python#security
Adversarial corpus: 50 destructive commands with expected tiers about 1 month ago
help wanted
empire-mind/agent-guardian empire-mind/agent-guardian A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Open project →
0
A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Python
#ai-agents#cli#devtools#python#security
Document 10 adversarial bypasses the regex engine misses about 1 month ago
good first issue
empire-mind/agent-guardian empire-mind/agent-guardian A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Open project →
0
A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Python
#ai-agents#cli#devtools#python#security
Add --version flag about 1 month ago
good first issue
empire-mind/agent-guardian empire-mind/agent-guardian A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Open project →
0
A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Python
#ai-agents#cli#devtools#python#security
Add ruff config and fix violations about 1 month ago
good first issue
empire-mind/agent-guardian empire-mind/agent-guardian A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Open project →
0
A one-file, stdlib-only firewall for AI agent shell commands: 0=clean, 1=WARN, 2=DENY — with machine-readable receipts.
Python
#ai-agents#cli#devtools#python#security