Plugins

Read-only behavior audit for DSH sessions: flags unsafe actions in 8 categories (A–H) with rules for 14 subclasses (the 41-subclass taxonomy file is not shipped, so a built-in 14-subclass fallback is used), anchors every finding to a real event seq, and gives deep links plus the full tool-call / data-access chain of any assistant reply (parameters, results, files read and written, domains, sandbox and approval context). Every finding is drill-downable.

Adds a DSH Settings panel for write-only local credentials and exposes only secret metadata and references to agents.

DeepSeek Harness tools to discover MCP configurations across clients, audit static security rules and probe a named server, using a separately installed mcp-hub Rust binary.

Keyword-rule scan of a project directory for tax-industry compliance gaps ahead of regulatory audits: accounting voucher and ledger retention set under 30 years, invoice-title lines with no taxpayer ID, tax data leaving China with no security assessment or standard contract mentioned, and e-invoice lines with no verification or de-duplication mentioned, with rule sources kept in rules/.

Grades every PowerShell command an agent is about to run before it runs: deletion, overwrite, network download and credential access are four required classes, plus dynamic execution, system and privilege changes. Verdicts are a pure deterministic rule function — no model call, no network, sub-millisecond — with two escalation rules (download+execute, credential+exfiltrate) that jump straight to high. High risk stops the tool call for a second confirmation (in-panel queue or the native approval), deny or timeout means it never runs; medium is logged but never blocked. Every command and its verdict, including the redacted command text, is appended to a local JSONL audit log. Quoted danger strings are treated as data, -WhatIf is respected, -EncodedCommand is decoded and analysed again. Host + client halves, zero runtime dependencies, 55 rules, 136 tests.

Capability drift detection: audit a plugin once, store a baseline of what it could do, then after every upgrade report only what is NEW (new install scripts, shell/eval access, outbound domains, credential paths). Deletions and version bumps are not reported, and no score is computed.

Device inspector and control panel for macOS and Windows that shows hardware, system, software, network, security and processes on one sidebar page and applies 19 whitelisted actions, with a confirmable cleanup scan, a repeatable benchmark and device_info, device_control and device_benchmark tools for the agent.

Turns every .pluginlet file under a repo's .dsh/pluginlets/ (directories configurable) into its own sandboxed Conversation View tab, gated by a capability grant bound to the file content hash.

Agent-asked secrets for DeepSeek Harness: a masked card stores API keys and tokens in the local credential store without them entering the chat, then injects them into one command.

Security review for DSH plugins: reads the manifest, install hooks, native dependencies and source for network egress, dangerous APIs and credential paths, cross-checks the curated market registry, and returns a 100-point verdict with itemised findings. Read-only — it never executes the audited code.

深鲸远程门禁：上线前扫描监听地址、鉴权开关、CORS、默认口令等远程暴露风险

一份能交出去的自查表、整改说明和合规证明

ArcVista 观界 —— DSH 内的架构观测台：本地项目、Git 克隆或批量 Git 导入，架构图、安全守护、知识图谱、复杂度与快照对比

Privacy guard for git workflows: before git commit, git push or gh repo sync it scans the content that would enter git (local paths, user info, API keys, private keys, connection strings) and blocks anything not approved in the DSH approval UI; approved items go into a manageable, masked whitelist. Native DSH Config schema with live settings, first-push full-repository scan, fail-closed checks, zero npm runtime dependencies.

Automatic approval and audit frame for DSH tool calls: a four-outcome rule table (deny, ask, AI review, allow) over commands, file access, network and tool scope, with security rules, rule-based auto-approve/auto-deny, AI fast/deep review, human approval, permission tiers, workspace boundaries, secret redaction and placeholder restore, a hash-chained audit trail and a simulator. Target files are inspected by metadata only.

检查日志是否开启、是否记录了该记的、是否留够法定 6 个月

按 GB/T 22239-2019 逐项对照代码与配置里可静态发现的等保项

DSH 插件：远程连接 + 逆向分析（REA / Ghidra）

检查告警链路：有没有告警、发到哪、有没有把敏感信息一起发出去

查应用里有没有展示 ICP/APP 备案号、隐私政策入口、运营者信息 —— 都是法定义务

查代码与配置里的境外数据出口：境外端点、个人信息外发、缺安全评估说明

审批台账 + 企业审计面板：记录每次审批（风险/决策/结果），gate 模式拦高危，Web 面板可看可导出

Code security defense review judge for DeepSeek Harness: detects sessions that are writing code, reviews them against the Code Security Defense Architecture Rules, recommends which security architectures the program needs (one or several), checks conflicts between them, teaches the session AI after repeated failures, and teaches the session AI after repeated failures; once you grant it review authority in its panel, it gates task completion until the review passes (without that authorization it only observes).

Local tool guard that scores every DeepSeek Harness tool call with an Ollama decision model before it runs; risky calls come back as an approval prompt, and an unreachable model stops all calls. Requires DSH 0.2.x and a local Ollama 0.35.1+ with the default ~6 GB clef-flash model pulled first; without them every tool call is denied.