Plugins

Cross-session memory for your dsh agent: remembered, recalled, and re-verified instead of assumed.

Agent 行为守护桌宠：跨 agent 监控可疑外传（加密打包直传对象存储 / DNS 外泄线索 / 敏感密钥被读），发现即告警，明确确认后按「一次一个」终止目标 agent；事件写入只追加 guard-events.jsonl，并通过 DSH host 工具 + MCP(stdio) 供其它 agent 查询

A clickable wooden-fish desktop pet for the DeepSeek Harness web GUI. Click to knock; counts per day with a calendar heatmap of your streaks in settings.

Conway's Game of Life for DeepSeek Harness: a sidebar panel row opens a drawable centre-column board with four rules, seven stampable patterns and a population chart.

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

Audits a DSH profile for capabilities nothing consumes: inventories the declared composition rows, counts how often each registered tool and skill is actually called in the session logs, separates calls that were made from calls that returned a result, and reports the ones with no observed consumer. Ships a skill that fixes the evidence-chain format for completion claims.

WeChat bridge for DeepSeek Harness over Tencent's iLink bot channel: one session per contact, inbound text/images/files, and WeChat's native typing indicator while a turn runs.

AlphaRSI — three-role self-improvement loop that evolves the training recipe of alpha sequence models. Operator runs walk-forward, Evaluator grades with OOS RankIC, Refiner proposes hyperparameters via TPE Bayesian search; only strictly-better recipes are kept. Offline one-command run, data bundle shipped with sha1 reconciliation, GPU auto-detect. Ships one skill.

DeepSeek Harness skill plugin that reviews law theses (international economic law and other legal topics): reads a student Word .docx, checks it against a bundled common-issue checklist, applies definite fixes as tracked changes, and adds Word comments for overall/section problems and missing-basis issues, then emits a short objective overall review plus a fix list.

End-of-chat visual storyboard: a storyboard tool renders a self-contained HTML card (inline SVG diagram, numbered flow, facts grid, timeline) from a JSON spec, writes it under ~/.dsh/storyboard/<session>/ and opens it in the sidebar; a system-prompt section and a /storyboard command trigger generation. Server-only, zero runtime dependencies.

Consistency guard for long-form writing: scans conversation prose for character, timeline and numeric setting conflicts across chapters, then jumps straight to the conflicting message.

9router-backed web search + fetch providers for DeepSeek Harness (ctx.web). Wires the native web_search/web_fetch tools to 9router's /v1/search and /v1/web/fetch endpoints.

Adds configurable HTTP header profiles that can be selected per model.

Adds a Global Prompt tab to the conversation view: text you write there is prepended to the top of every session system prompt, and saving takes effect on the next message without a restart.

Delete a complete closed turn (question + answer + tool calls) from a DeepSeek Harness session, without deleting the session.

Maintains stable emoji session titles in DeepSeek Harness after user prompts without adding naming messages.

Mirrors cc-switch providers and models into DeepSeek Harness the moment they are saved, and shows each provider's plan windows, account balance, or custom usage script in the composer and the settings card.

Always-on-top lamp board for macOS: one lamp per agent, one row per session, collapsing to one lamp per running task; a lamp turns blue and blinks while an approval or a question is waiting on you, and the root agent finishing raises a system notification.

Adds a Restart page to the DSH Web settings panel. The page shows the PID, port, start time and uptime of the running host and restarts it through a detached helper that waits until the port stops answering before replaying the same launch invocation; the page reloads itself once the replacement host answers with a new boot id.

A draggable bubble that picks three dishes at a time from a local 4,423-dish library; the first two are always well-known mains, and picking costs no tokens.

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.