返回目录

GITHUB TOPIC

guardrails

7个项目包含此标签

7 个项目

GitHub Topic 精确匹配

安全与治理

⚡ Millisecond System-1 judgement for every tool call in DeepSeek Harness — Jev-powered risk classification & evidence-gated auto-approval. Fail-closed by construction. dsh 生态第一个 System-1 决策插件

已验证
agent-harnessagentic-aiai-agentsai-safety
开发工具

SpecWave

Cyning12

SpecWave — multi-host coding CLI + P0 gates/Harness (Cursor/Claude/DSH). Formerly SpecGate / dsh-coding-kit. npx spec-wave

已验证
agents-mdai-codingclaude-codecli
安全与治理

dsh-safety-net

Asuna486-desuwa

给自主运行的 Coding Agent 加的自我保护护栏:拦截对受保护路径的写入、破坏性操作前自动备份、CLI 自恢复命令

已验证
deepseek-harnessdsh-pluginguardrailssafety
Agent 与会话

DeepSeek Harness plugin: a code-level guard that resolves post-task confirmation items by weighing output quality against real user impact.

已验证
agent-pluginsai-agentscordiscordis-plugin
安全与治理

dsh-rule-lens

jackeyunjie

规则透镜(RuleScope Lens)— DSH 本地插件:AGENTS.md/rules 加载可视化、预算治理、硬拦截、Lint、遵守率 | Rule visibility, budget governance & hard guards for DeepSeek Harness

已验证
agents-mdcontext-engineeringdeepseek-harnessdsh-plugin
开发工具

dsh-change-budget

Raphaelutumn

DeepSeek Harness file-safety guardrail that limits AI coding-agent file edits per turn and prevents too many file changes, with call and UTF-8 byte budgets. 限制 DSH Agent 每轮修改文件数量。

已验证
agent-guardrailsagent-safetyai-agentai-coding-agent
开发工具

Jev (TypeSafe System One) for DeepSeek Harness: five verdict tools - jev_decision, jev_choose, jev_overview, jev_guard_status, jev_verify - with in-chat views, real system-prompt guidance, a full settings card, an auto-guard (deterministic + Jev risk/loop) and a /jev dashboard. Real API only, measured numbers.

已验证
ai-agentsdecision-modeldeepseek-harnessdsh