Seguridad y permisos

Higiene de texto como plugin de dsh: sanea texto no confiable, detecta caracteres invisibles, limpia el formato generado por LLM y escapa la inyección de fórmulas CSV.

Escalation-only LLM reviewer: reviews sandbox escapes only, keeps the sandbox, fails closed.

中瑞世联工作台 / CRWU audit workbench for DeepSeek Harness: pick a pending audit report, dispatch one AI audit subagent, watch and stop/restart it, auto-upload deliverables to Aliyun OSS.

dsh plugin: unattended approval gate — subagent/scheduled/cron approval requests get a bounded answer window, then settle with a fail-closed default instead of hanging forever (per origin or per session)

Harness-neutral Traditional Chinese enforcer + offline converter (skill, DSH bundle, CLI). One script axis either way + Cantonese/Japanese-only filters. Converts both ways; optional wording preference. Also: an LLM output guard, cp950/GBK console scan.

A structured decision layer for the DeepSeek Harness agent loop: a danger-call gate, output self-check, tool narrowing, and loop guard backed by a pluggable adjudication model.

Make DSH llm-pi-ai opencode routes pass Zen free-tier gate: opencode User-Agent + canonical ses_ session + bash/read tools. Scoped to opencode.ai only.

Host-only command safety gate for DeepSeek Harness: blocks recognized dangerous operations and requests DSH approval when a command cannot be inspected.

Audit DSH skills automatically: a host-layer tools/post-execute plugin that ships both the skill body and the audit engine, runs the engine after skill files change (write/edit/shell) or after a bulk restore/backup (any tool invoked with mode: restore|bac

Pure JavaScript macOS Keychain credentials provider for DeepSeek Harness, replacing plaintext .credentials.yaml storage and fully supporting both refs and records without native builds, signing, or Xcode.

Shared infrastructure for data assetization plugins: business rules loader, file format adapter, report generator, path validator, audit logger Also known as @deepseek-ai/dsh-data-asset-shared.

Approval-free writes into whitelisted directories — wraps the filesystem fence for write/edit tools and auto-answers bash sandbox escalations that touch whitelisted paths, with a settings GUI.

Contract check for DeepSeek Harness plugins: statically verifies that registered tools' output.render() returns ContentBlock[], and warns when a session event type falls outside the kernel vocabulary. Warn-only. 契约体检插件（不检测恶意）。

Host- and agent-layer DeepSeek Harness plugin that prices every automatic context rewrite before it lands, asks the human by default, and reports what the provider had to re-read cold.

Maintained fork of dsh-approval-gate. A neutral operation whose confirmation count has reached the threshold auto-approves when the judge is unavailable, instead of prompting a human again; approver-facing explanations are generated in Chinese from the real sandbox mode, command and paths. Also carries deterministic hard-deny for credential exfiltration and system-path destruction, judge-input redaction, a judge model candidate chain, fingerprint-scoped allow rules, and an approval view with unified diff and one-click revert.

Pre-execution safety valve for DSH that judges with TypeSafe Jev: it hooks tools/pre-execute and decides every bash/pwsh call by first applying offline static rules, then by asking the Jev model — the System One model from TypeSafe, which returns a structured decision instead of prose — one yes-no question ("will this command irreversibly delete or overwrite real data?"), turning the answer into allow, revise, block or escalate; revise hands the model a safer rewrite, escalate offers a one-shot human token; exhausted credit or a missing API key degrades loudly instead of failing silent, and a missing key is requested in the conversation itself (`guard key set`, read from stdin); every verdict is appended to a shared audit log and the text people read is bilingual zh-CN/en.

Customize the dsh UI theme: 12 accent presets, dark-mode aware accents with a contrast guard, density, font family, animation toggle, and raw design-token overrides.

Zero-intervention permission gate for DSH agents on macOS: auto-allows local writes made reversible by git or Time Machine, blocks network, package installs, process control and sensitive-path reads, with full audit logs.

让 dsh 在 Windows 上使用用户自安装的 Git Bash，并绕过其沙箱限制 | Enable dsh to use a user-installed Git Bash on Windows by bypassing its sandbox restrictions.

One-click Approve & escalate on the approval card: answers the pending request and switches the live session to a permission preset.

Query and manage JumpServer through conversation: assets, users, accounts, permissions, sessions, command audit logs, command filters, and RBAC roles, authenticated with an AccessKeyID/AccessKeySecret pair (HTTP Signature).

Approval-free writes into whitelisted directories — wraps the filesystem fence for write/edit tools and auto-answers bash sandbox escalations that touch whitelisted paths, with a settings GUI.

A continuation of simon300000/dsh-auto: the Auto Approve permission preset for the DSH Web UI reviews each approval with one single-shot model call instead of a reviewer subagent, auto-approves only the actions that pass and hands the rest to human approval, and remembers confirmed decisions as project- and global-scope allow/deny lists, with an approval timeline in the right sidebar.

Keeps a redundant sandbox-escalation argument from failing a tool call. Escalating tools (pwsh, bash, write, edit) advertise the full sandbox_permissions enum, but DSH only accepts a level strictly wider than the one already in effect — a check its own source calls 'deliberately not a schema constraint'. A model that reflexively passes the argument therefore picks the level it is already at and the call dies before running with 'sandbox escalation to "workspace-write" is not strictly wider than this call's current "workspace-write" mode', costing some models a whole turn of retries. This plugin registers one tools/execute waterfall listener and, only on that documented rejection and only when the arguments really carried an escalation field, re-issues the identical call once without it. Safe by DSH's own documentation: the rejection precedes any execution ('nothing has run'), and the corrected call cannot match again, so the retry is loop-free. Reproduced and verified end to end — before: isError true, the command never ran; after: the command's real output, isError false, one tool/call and one tool/result. Zero dependencies.