安全与治理
插件
Second-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent returns structured allow/deny verdicts with reasons, fail-closed by default, fully auditable from the session log (approval/asked -> autoReview/verdict -> approval/decided).
ai-safety
approval
auto-review
cordis
安全与治理
插件
Fail-closed export-copy redaction for DeepSeek Harness session telemetry
deepseek-harness
dsh-plugin
privacy
redaction
安全与治理
插件
Deterministic fail-closed tool-call authorization for DSH with evidence: allow/block/ask policy gate plus approval-chain deferral.
agent-safety
agent-tooling
ai-safety
deepseek-harness
安全与治理
插件
dsh-yolo-mode - an LLM-powered auto-approval plugin for DeepSeek Harness sandbox escalations (built-in presets + custom permission levels, fail-closed)
approval
deepseek-harness
dsh
dsh-plugin
安全与治理
插件
AI approval answerer for DeepSeek Harness (DSH) using the unified LLM route with fail-closed policy checks.
agent-harness
approval
cordis
deepseek-harness
安全与治理
插件
Local, zero-network, fail-closed secret preflight for DeepSeek Harness model requests.
deepseek-harness
dsh-plugin
llm-security
secret-scanning
