fu351/Doberman-Core

QA: thoroughly test everything merged to main after 2026-06-20 (#18, #21, #26–#29) about 23 hours ago

help wanted

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

Review request: PR #47 (red-team gate ratchet + channel coverage) — ready, needs a peer to merge 2 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

Review request: PR #54 — Windows onboarding crash fix (encode-safe CLI output) 4 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

Review request: host-hooks stack HK.1-HK.5.1 (PRs #32->#44) — all ready, needs a peer to review + merge 4 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

hosthooks: PreToolUse cwd="." fallback may load the wrong strictness mode (HK.1) 4 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

tests(hosthooks): isolate test_interactive_global_flag_overrides_prompt — it writes to the real ~/.claude/settings.json 4 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

rule-lab: harden the acceptance gate — bound --eps-fpr and enforce tune/holdout disjointness 4 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

tests(ood): add red-team corpus entries for invisible_only and unpaired_surrogate channels 4 days ago

help wanted

fu351/Doberman-Core

111

Doberman is an AI agent security framework for guardrails, prompt injection defense, runtime policy enforcement, tool-use permissions, agent monitoring, audit logs, LLM safety, autonomous workflow protection and secure AI deployment.

Python

Open Issues Need Help