Plugins

MCP server management for DeepSeek Harness: author composition rows, choose when an allowed server loads, and filter which of its tools a session may call

A newspaper-style front page for a session you left running, as a Conversation View tab beside Chat and Trajectory: a per-turn timeline (duration, steps, tools, tokens, outcome), file churn counted per file, where the wall clock went and how much of it was the model rather than tools, which single turn ate the spend, failures nobody retried quoted with their failing command, and a context-pressure gauge measured against the routed model's own declared capacity. The briefing is folded from the durable session log and never generated by a model, so it costs no tokens, cannot hallucinate, and is byte-identical for the same log and clock; it appends nothing to the session and reaches no model, and it reports a pending approval without ever answering it. Pages are offered only when they have content, and the tab refreshes itself on turn settle and on a two-speed poll backed by a metadata-only tail probe.

DeepSeek Harness plugin that asks which provider and model should run each delegation before the child starts, so a subagent can go to a cheap or local route instead of the parent's.

A per-turn change bar above the composer showing which files the turn changed and the line deltas, for DeepSeek Harness 0.1.5-rc.1 through 0.1.7.

Fixes the Windows workspace-open action. The shipped opener runs `explorer.exe "<file:// URI>"`, which on some hosts reports success without raising a window — the exit-code verdict cannot tell an opened window from nothing happening. This routes that one action to `explorer.exe /e,<dir>`, which raises a window reliably, and leaves every other application in the menu on the shipped path.

Maintained fork of feibi-mochi/deepseek-harness-control-center, which has not been updated since 2026-09-09. Pricing follows the official model pool instead of a hardcoded model prefix, so DeepSeek V4.1 Flash (`deepseek-flash`) is billed and shows its peak/off-peak clock, and the official price page is parsed structurally so a renamed or retired model no longer stops price sync. Adds an English interface beside the original Chinese. Otherwise the same plugin: balance, per-session cost and token breakdown, a 24-hour peak/off-peak clock, Z.ai plan quota, and a local usage ledger, following the host light/dark theme.

Rapid-refill circuit breaker for DeepSeek Harness automatic compaction. It refuses a compaction that would be futile — the context refilled within fewer than N tool turns M times in a row, which the unbounded step-pressure trigger cannot detect — and instead of burning a summarization call on every step it reports the situation with advice, usually that a single read or tool output was too large. Shows state and re-arms through /compaction-breaker, and injects one prompt section so the model relays it to the user. Covers host-plane sessions through its profile bundle; ordinary sessions need one row replaced inside an agent preset.

Live progress for long-running DSH tasks: a script appends JSON lines to a per-session progress file, the Web UI draws them in a floating overlay and a right-sidebar tab that also lists background jobs which reported nothing, and `dsh-progress run -- <cmd>` wraps a command that cannot report for itself.

Whale-girl desk pet in the corner of the DSH Web GUI. It sings along to a bundled track while the agent works; when a tool call fails a steel pipe falls on its head. Click it to bonk the basin it wears, which costs 12 rice — knock enough and it gets hungry and wanders off to beg, and rice appears somewhere on the page for it to dash across and eat. Draggable within the viewport, wriggles and complains if grabbed mid-dash, volume and behaviour configurable from Settings. All artwork, sound effects and the song ship inside the package.

Ships a skill about handing work to subagents instead of doing it inline - which tasks are worth fanning out, how many to run at once, how to send them in parallel, and how to check what they returned instead of trusting the summary.

Browse the awesome-dsh-plugin catalog from a page in DSH Settings: filter by category or by installed state, search by name or description, sort by stars, downloads, name or date added, and install a plugin into the current profile. Two further tabs list what is already installed and which of those have a newer version in the catalog, with an update button on each row. The update view reports installed plugins whose update status cannot be determined, and overlapping install requests are refused. The catalog is cached locally for an hour with a mirror, a CDN and the npm registry as fallbacks when the official source is unreachable.

Ships the Zhouli-mode agent preset, which runs a session in three phases (declaring roles on opening, citing authorities while working, a six-item self-audit at the close), plus one tool with seven actions over a bundled 49,384-character 《周礼》 corpus that returns every passage with a stable 篇序.职官序 coordinate, a per-office establishment index and collation between two Kanripo witnesses.

Explains which layer killed a failed turn: times every model request (first byte, longest silence, total) and combines that with the error to name the culprit (dsh idle watchdog vs Node undici timers vs SDK timer vs server crash, context overflow, gate queue, tool timeouts, failed compactions) and the exact setting to change; verdicts in the DSH terminal plus a /why command, subagents included.

Context-window attribution for DSH Web: a Mini Utility Dock panel that ranks the user messages, assistant turns, reasoning blocks and tool results of a session by the tokens each one occupies, so it is clear which entries hold the window. It also shows the share by entry type, the heaviest entries across the live sessions of the host, and, where the host provides them, provider usage and the estimated system/tools/messages composition. The panel is read-only: it does not estimate cost, modify a session, or trigger compaction.

A /remember command that writes facts into the AGENTS.md instruction hierarchy dsh already loads, plus a Memory tab showing which instructions the loader kept within its byte budget.

Turns attached files into versioned evidence: `search_documents` builds a private local index (SQLite FTS5 after a startup capability probe, dependency-free JS fallback otherwise) and returns compact evidence blocks carrying an exact coordinate — PDF page, PPTX slide, text/DOCX line range, or quoted XLSX `Sheet!Range` — which `read_document` expands only while the content version still matches. Contiguous CJK runs are indexed as overlapping bigrams and queried as phrases, so word order is preserved; uploaded raster images take the native vision attachment path instead.

Tells the dsh-desktop app which project the open session works in, and puts files it sends into the chat

Control which tool schemas are injected into the DeepSeek Harness model context. Registers a tool_visibility manager tool AND a settings-page UI (Settings → 工具注入控制) with per-tool toggles; backed by the framework-native tools.restrict filter and persisted

Live MCP server status in the DeepSeek Harness conversation composer — see which MCP servers are actually connected before you hit send.

MUV engine for DeepSeek Harness: runs the regex_scripts from a character card against model output, tracks variables, renders a status bar, expands macros and LaTeX, and styles tagged code blocks such as letter, terminal, newspaper and phone. Needs dsh-muv-table installed as well, which is its companion plugin.

Memory triage for DeepSeek Harness: at the end of each turn it decides which entries to keep and which to archive, routes project-specific ones into a project bucket while global ones stay in the main store, merges near-duplicate entries, and falls back to an LLM when the rules are unsure — automatic actions are recoverable archives only, never deletions.

A DeepSeek Harness plugin that records which files the agent edited, with which model, and in which session into git-ai

A DeepSeek Harness plugin that records which files the agent edited, with which model, and in which session into git-ai

Reports the semantic blast radius of agent code edits: which symbols changed, who calls them, and which of those callers have no covering test.