@deepseek-ai/dsh-sandbox-microsandbox
Fail-closed microsandbox microVM provider and model-facing guest tools
122 results
Fail-closed microsandbox microVM provider and model-facing guest tools
Allow, ask, or deny DeepSeek Harness tool calls before execution
Automode for DeepSeek Harness: a fourth permission preset that runs on full access with an LLM classifier as the only gate before every tool call
Audit deepseek-harness (DSH) plugins before install, guard them at runtime. Static verdict + pre-install audit protocol; supply-chain checks (typosquat, OSV); exfiltration & ransomware detection, honeypot canaries, integrity baseline. Alarm-only; blocks confirmed destructive ops.
🐳 The Everything-Plugin Quant OS — AI-native & DSH-native: 46 tools, 6 pluggable domains (data/alpha/ML/risk/execution), one end-to-end PDAT→PET pipeline. Methods open, secrets internal.
DSH plugin: strip GPT-family tool-call sandbox_permissions that are not strictly wider than the current session.
工具调用规范守卫:对 agent 工具调用参数做字符串匹配,命中危险行为拦截并注入原因(deny),或放行但注入警告(warn),附规则管理面板
Model-facing typed tools wrapping the frozen, reviewed underseal adapter for the DeepSeek Harness
DeepSeek Harness plugin (temporary, pre-upstream-fix): automatically injects a CPython tempfile shim (PYTHONPATH -> sitecustomize) into every confined shell command on the Windows sandbox, so python/pytest tempfile use works with zero extra tools, zero model-context overhead, and zero escalation.
工具调用规范守卫:对 agent 工具调用参数做字符串匹配,命中危险行为拦截并注入原因(deny),或放行但注入警告(warn),附规则管理面板
OMV Audit Desk: an evidence-first vulnerability audit workbench for DeepSeek Harness
自然语言查询 + 商业智能报表插件 for DeepSeek Harness:在 dsh-mysql 连接底座之上,提供 Text2SQL 自然语言查库、15+ 图表类型、Dashboard 仪表盘、指标/维度管理、多维分析、自助分析、报表导出、审计日志与数据权限。Natural-language data queries + enterprise BI reports for DeepSeek Harness: Text2SQL, 15+ chart types, dashboards, metric management, multi-dimensional analysis, self-service analysis, export, audit log and data permissions.
DSH WebUI capability controls for skills, MCP servers, tools, prompt injections, approvals, and safety guards at session, project, and global levels.
Hard-stop consecutive identical DeepSeek Harness tool calls after a configurable streak.
DeepSeek Harness 工作区管理工具集:查看、创建注册、重命名、删除工作区记录(纯注册,不触碰磁盘目录);破坏性操作(删除注册、改注册路径)走用户审批 fail-closed。
Audit an agent harness against the harness-evaluation criteria, with machine-enforced evidence validation.
A runtime prompt-injection shield for DeepSeek Harness tool results.
V0 capability-disclosure scanner for DeepSeek Harness (DSH) cordis plugins: reports which sensitive services (credentials/shell/approval/sandbox/fs/tools...) a plugin declares or references before you trust it.
Mingleng mcpguard for DeepSeek Harness 鈥?the first security plugin for DSH. Scans skills and MCP configs for prompt injection, homoglyphs, hidden Unicode, dangerous shell and credential leaks.
Secret & dangerous-pattern scanner for DeepSeek Harness — a security_scan tool that finds leaked API keys, tokens, private keys and credential files, with full redaction.
Autonomous (auto) mode permission classifier for DeepSeek Harness: a Claude-Code-auto-mode-like classifier over tools/pre-execute and approval/request, a selectable 'auto' permission preset, LLM semantic judge, git checkpointing, agent discipline guidance, and a web control page in Settings → Plugins.
Fail-closed DSH compatibility guard for redundant GPT/Codex sandbox escalation arguments
Turn DeepSeek Harness sessions into redacted, README-ready animated demos — as a local CLI or DSH plugin.
Strip sandbox_permissions / justification from model tool-call arguments so sandbox escalation is never triggered when the session already has sufficient permission.