347 results
为 DeepSeek Harness 增加介于 Workspace Write 与 Full access 之间的自动批准权限档：例行沙箱升级由分类模型一次性放行，危险或不确定的操作仍转人工审批。An auto-approval permission preset for DeepSeek Harness between workspace-write and full access: routine sandbox escalations are granted once by a classifier, while dangerous or uncertain requests still go to a human.
DSH local security audit tool: read-only scans of config, credential metadata, plugin provenance, session structure and network exposure; redacted, reproducible, locatable risk reports
Private, zero-trust gateway for DeepSeek Harness Web: loopback-only proxy, exact principal allowlists, and provider-agnostic onboarding
DeepSeek Harness 远程与移动端安全网关插件：零修改 DSH 底层代码安全开放局域网与 Tailscale 连接 | DeepSeek Harness (DSH) Remote & Mobile Security Guard: safely opens Tailscale/LAN with zero core modifications, QR scan auth, RSA encryption & brute-force defense.
DSH plugin: bind the Web GUI to 0.0.0.0 for LAN access and polyfill crypto.randomUUID for non-secure (LAN HTTP) contexts. Trusted networks only.
Per-model capability declaration for DeepSeek Harness: reasoning-effort levels (with wire spellings) and request modalities (text/image) for OpenAI-compatible providers — one settings section, no YAML hand-editing.
Fail-closed LLM-assisted approval reviewer for DeepSeek Harness
Render ```mermaid fences in DSH Web conversation messages as interactive-free SVG diagrams (lazy-loaded, strict security).
Security scanner for DSH plugins: static and semantic passes over plugin source, DSH-specific attack-surface rules, npm audit, batch scanning, and an HTML report with per-finding severity and evidence.
Password login gateway plugin for dsh web: auto-generated initial password with forced onboarding, login, password change with real request interception, and OTP two-factor authentication.
DSH plugin exposing a single `sleep` tool that lets the model pause for a configurable number of milliseconds, honoring cancellation.
DeepSeek Harness plugin: secondary working directories for a project. The agent keeps the primary workspace as cwd, gains equal write/exec permissions on configured secondary directories under workspace-write mode, and is notified of configuration changes at the next message boundary. Configurable from the session header AND from the session-creation page (before the first message) through a sessionless multiFolder remote API.
Codex desktop-pet skins migrated into DSH: auto-detects community Codex pet skins, copies them into the DSH home, and renders an animated desktop pet with per-session status dialogs and a settings section.
DSH 插件：全模式工具失败自动实录——把原生工具、PTC(Code Mode) run_code、代码内嵌工具调用的失败按错因去重计数，自动写入 skill 的机器维护区段，越用越少错。Auto-record tool failures from every execution mode into a skill's machine-maintained section (dedup + counts).
SummerSec personal Agent Skills monorepo and DeepSeek Harness profile bundle; version mirrors multi-platform plugin manifests.
Persistent long-term persona memory for DeepSeek Harness: MEMORY.md/USER.md storage, memory + memory_search tools, per-request context injection, secret scanning, and a WebUI memory-management settings page. Ported in spirit from pi-hermes-memory. TypeScript sources, built with tsc + esbuild.
Modular AI research/engineering skill pack for DeepSeek Harness — security audit, paper workflows, dev workflow, storage analysis. Installable via dsh plugin add.
Config-driven lifecycle hooks plugin for DeepSeek Harness: declare event -> command hooks in cordis.patch.yml, no plugin code required. Includes a Hooks section in the Web GUI settings (history timeline + manual tester + notify tests + hook editor + Feishu connect).
Injects one system-prompt section so a DeepSeek Harness agent thinks and replies in Chinese
Health checks for DeepSeek Harness: manifest/patch/entry/build/pack/install verification, model-callable plugin_check, profile host-shadowing + manifest-BOM detection, environment diagnostics, supply-chain poison preflight, and broken tool-call-sequence detection (#2334).
Brings open-mcp-apps into DeepSeek Harness (dsh): an Apps section in the sidebar, one container per app with its own workspace and conversation, an agent-presence strip under the app, and inline app rendering for MCP tool calls.
DeepSeek Harness (dsh) plugin: run work as a small crew of role agents (product manager, researcher, architect, engineer, test engineer, code engineer, QA, code reviewer, security reviewer, doc reviewer) that talk through files on disk, with the PM as the only voice to the user.
Merge consecutive same-tool calls in the DSH WebUI chat flow into one card with compact child rows (all built-in tools by default).
Detect which conversation model your DeepSeek Harness account is served: score probe conversations' live chain-of-thought, cancel old-model probes in seconds, keep confident catches.