dsh-strip-sandbox-permissions
Strip sandbox_permissions / justification from model tool-call arguments so sandbox escalation is never triggered when the session already has sufficient permission.
191 results
Strip sandbox_permissions / justification from model tool-call arguments so sandbox escalation is never triggered when the session already has sufficient permission.
Database connector bundle for DeepSeek Harness (dsh): SQLite/PostgreSQL/MySQL connections, schema introspection, read-only safety, write approval gate, and a JSONL SQL audit trail.
Runtime security guard for DeepSeek Harness (DSH) — blocks command injection, SSRF, credential exfiltration, and destructive operations at runtime.
DeepSeek Harness GitHub plugin: review PRs, triage issues, debug Actions CI, handle review feedback, and prepare code changes for review — via gh CLI or REST, no extra sign-in required.
DSH 数据库 MCP 桥接插件:零知识凭据管理(密码永不进模型)+ 按 source 的一次性进程执行(无常驻服务、天然并发与多实例安全)+ 自动安装 dbhub 与浏览器状态卡片。
DeepSeek Harness (DSH) host-layer plugin: detects and auto-fixes Windows PowerShell command syntax mistakes — bash constructs, broken line continuations, pasted prompts, CRLF pollution — then executes the corrected command through the host shell seam under the normal sandbox/approval policy
Recover deleted/corrupted dsh sessions (session.jsonl.zstd) and memory (memory.db) straight from the raw disk. Battle-tested recovery manual + scripts + a /session-repair command plugin for the dsh web UI.
Interactive 3D molecule viewer for DeepSeek Harness: view PDB/SDF/MOL2/MOL structures with 3Dmol.js
Local-first RAG knowledge tools for DSH: kb_query / kb_ingest / kb_crawl / kb_related, backed by the kb-rag Python pipeline (Ollama bge-m3 + ChromaDB + SiYuan). Zero API cost, data never leaves your machine.
DeepSeek Harness 插件:解析 Word/PDF 题目,支持本地智能索引/聚类/检索/抽题,生成带书写空隙的可打印试卷
Workspace Only switch for DeepSeek Harness: confine file access of every sandbox mode to the session workspace
Guardian state engine for the DeepSeek Harness — assertion counting, edit review queue, and no-feedback-loop monitoring over session events
Real-time code feedback for DeepSeek Harness — LSP, linters, formatters, type-checking, and structural analysis.
Seatbelt (libsandbox) sandbox provider for the DeepSeek Harness sandbox seam on macOS: a cordis plugin that replaces @deepseek-ai/dsh-sandbox-local's seatbelt rung via a prebuilt seatbelt-run launcher — no dsh source changes, no PR required
Self-protection guardrails for the DeepSeek Harness: protected-path interception, backup-before-destroy, CLI self-recovery commands, and strict sandbox defaulting
One-click control that closes a session's persistent bash terminal (dual-face: /exit-bash command + session-header button) so the sandbox mode can be switched
DSH web plugin: play a notification chime when the agent asks the user a question (ask_user_question / plan review) or requests approval (sandbox escalation, tool approval).
DSH 桌面通知 — 任务完成 / 等待输入 / 审批被拒 / 子代理与后台任务结束时推送原生桌面通知(Windows 用 koffi 直调 WinRT,Linux 直连 D-Bus,无 Python、无子进程);仅 DSH 网页聚焦时静默,非聚焦推送。Installed as a real web-profile plugin: auto-loads on every dsh web start, no approval needed.
DSH microsandbox bundle: Linux microVM shell executor and ctx.sandbox provider (msb run) with workdir/env preservation, fail-closed probing, and shell-free argv construction
Stent MXC sandbox sidecar for the DeepSeek Harness
DeepSeek Harness bundle that routes real sandbox escalation prompts through a dedicated approval reviewer
dsh-langfuse: Langfuse LLM observability for DeepSeek Harness — one OpenTelemetry trace tree per agent session, with feedback scores and subagent lineage.
Tiered auto-review for DeepSeek Harness (DSH): static-rule safety net + LLM reviewer + human fallback. Auto-approve safe actions, auto-deny irreversible ones, ask a human for the rest. PURE VIBE CODING - not audited, use at your own risk.
Better config hub for DeepSeek Harness: one page showing sandbox mode, default model, workspaces, settings namespaces and copy-ready composition templates.