Plugins

Adds a Graph tab that renders one session's execution as an interactive DAG of turns, LLM steps, tools, skills, subagents, code changes and tests, with live status, retry aggregation, critical path and trajectory linkage.

Rewind / retry / fork orchestration (host) plus conversation.chat.node renderer replacement (browser) — message edit / retry actions on user bubbles.

When a dsh turn gets interrupted — by a web restart or an upstream hiccup (rate limit, 5xx, quota error) — this plugin auto-injects a "continue" so your session picks up where it left off, with a dead-loop guard that hands back control if the retry produces nothing.

Turn-atomic message editing, retry, reroll and version-tree navigation for DeepSeek Harness conversations

Anti-bot HTTP fetch tools hardened by SXS's production scraping stack (millions of requests/day): UA-pool rotation, retry with exponential backoff, anti-bot wall detection (captcha / verification challenges) and adaptive rate limiting — tools: `sxs_fetch` / `sxs_fetch_json` / `sxs_rate_status`.

Undermind literature search inside DeepSeek Harness. The plugin owns the OAuth authorization-code + PKCE login the remote MCP server requires, refreshes the rotating refresh token before expiry, re-initializes the MCP session when the server drops it, and registers every tool Undermind tools/list returns as a mcp__undermind__* tool, alongside a permanent undermind_auth tool for status, login, logout and retry. Needs an Undermind account and a client id - Undermind accepts a Client ID Metadata Document URL (CIMD) or a pre-registered client id, and does not support dynamic client registration.

Stops the agent after N consecutive failures on the same tool. Instead of another retry it pushes a four-part stop instruction: what was tried, the exact error text, the likely cause, and what the user must do. Detection reads real tool results (isError or a non-zero shell exit code) and mirrors them through agent.steer, so the agent sees it as an inbound message rather than a log line.

dsh plugin: tool failure radar — error rates, retry clusters and slow tail per tool from the tool-trace sidecars.

Transparent cross-provider failover for DSH Agent main requests: named fallback chains, per-session chain choice, per-route retry overrides.

Multi-role group chat panel for DeepSeek Harness: join as one character, chat with 2-3 AI members, route replies by addressee with probabilistic bystanders, use @ handoffs, and keep private memory/relationships, a global character library, failure diagnostics, retry of failed replies, message search, a sticky scene note, paged long-history rendering, cross-tab sync, provider-catalog models, and JSON import/export.

Forensic analysis of past model-routing decisions. Replays each session, classifies every turn's task complexity from its request shape (deterministic, no LLM), and reports two kinds of waste: an expensive model on cheap work (the difference is measurable), and a cheap model on expensive work that failed on its own merits — where every retry re-sends the whole context, so the downgrade cost more than the upgrade. Distinguishes provider-capacity retries (RATE_LIMIT/TIMEOUT/TRANSPORT) from model failures so it does not cry wolf. Derives routing rules from the evidence, priced against a built-in per-million table.

Automatic recovery when a turn ends badly: resume after a transient failure, continue after output truncation, re-run after an unfinished todo list, plus a request-level 429 backstop and a client dock with edit-resend, manual retry and stop-and-re-ask.

Token usage dashboard for the whole harness: billed tokens from every session are rolled up per session and turn (uncached input, cache read, cache write and output) with a peak/off-peak CNY cost estimate, and a "Token 用量" entry above Settings in the sidebar opens the dashboard at /token-stats, in a new tab on the web and in an in-app popup on the Desktop. The page recollects and repaints every 60 seconds while the tab is visible. Usage is read from the same session event stream dsh-token-meter consumes, accepting 0.1.x assistant/chunk and 0.2.x assistant/message and assistant/attempt, and counting an attempt repeated after an llm retry as its own billed call; cold history is seeded once from a persisted rollup file and afterwards only logs whose mtime or size changed are re-parsed. Host half and client half ship in the one package.

Read-only clinic for DSH session logs: 22 named diagnostics (token waste, retry and failure loops, context bloat, cross-session amnesia, idle MCP servers and skills), each reported with a replayable evidence locator such as session.v4.jsonl.zstd#L101-L110, and each finding turned into an executable prescription — a profile-patch YAML snippet or a command with current to suggested values, risk, verification and source — plus a feedback command that retires a prescription you reject, after which the report states it was removed by your feedback. Zero dependencies, default zero egress, session files are never written.

A prompt optimizer in the composer input tool row: pick a category and a template, then ✨ rewrites the draft in place with one-click undo. Pick from the Basic, Context and Image categories; you can cancel a run, and it shows elapsed time and token usage. Ships 15 templates covering task-instruction rewriting, step planning, safety-research framing, system-prompt role cards, and text-to-image / image-to-image prompts; context templates read the recent turns of the session for coreference resolution. Calls go through the host model service, and a settings section exposes the model and the budgets. Output guards check that user-written constraints survive, warning rather than auto-retrying. Adapted for dsh 0.1.7 and 0.2.

Per-project MCP servers: a project declares servers in its own documents (`.dsh/mcp.json` by default; the local and global file lists are settings), and they are mounted only for the sessions working in that project — on the first turn of a session, released when it goes idle. A right-sidebar tab and a Settings page list each project with a live session, the status of every server and the session holding each mount, with `Sync` / `Retry` / `Release`, an `enabled` switch and an editor over one existing entry; a per-project tool policy (`disclosure` / `direct` / `off`) with pins and `mcp_search_tools` decides what the model sees; `${VAR}` references resolve through the project `.env` files and the credentials file; per-server call counters outlive the sessions that produced them.

Model channel manager: several provider/model candidates share one virtual route (roundrobin/<group>) with failover, retry, timeout and cooldown; a settings tab edits providers and groups, runs per-model speed tests and real single-model request tests, and a seven-day health panel aggregates latency, failures and upstream token usage (input, output, cache and reasoning).

Image generation workbench and agent tool for DSH: one tool (ark9_generate_image) paints text-to-image and image-to-image via any OpenAI-compatible API. Channels are entirely user-configured - no bundled endpoint, you bring your own baseURL, key and model name (manual input supported). Agent-initiated generations are approval-gated by default: requests queue in the floating panel and nothing bills until you approve; approval is non-blocking (the agent replies "awaiting your approval" immediately, and results, denials or timeouts are injected back so the agent reports them in the next turn). Floating glass workbench with five tabs (Generate / Approvals / Prompts / History / About), aspect-ratio grid with quality-budget + 16px-alignment sizing, transparent background, batch up to 10 images as aggregated sub-tasks, prompt library with custom JSON sources fetched through a host proxy, persistent generation logs with one-click retry, multi-channel aggregation with per-channel model fetching, bilingual UI (zh/en), stroke-SVG icon set, config import/export. Registers as a Better Sidebar tab when dsh-better-sidebar is installed; stacks above the dsh-cua FAB when both are present.

Session health-check: mechanically mines session logs into structured problem tickets (retry loops, tool-error clusters, interrupted turns, token waste) over the official session-query seam; works standalone.

Provider-first three-level model selector for the DSH Web model seat: pick a provider, then a model from that provider, with inline search on both levels, failed-provider retry rows, and a Model · Provider trigger.

Provider/model-aware adaptive pacing and visible rate-limit retries for DeepSeek Harness free models.

A permanent DSH plugin that bundles a developer workbench, an in-session conversation tree, and a plugin manager: build and run an LVAL project, browse its code, and keep per-program version snapshots with rollback; edit, retry, or branch any message in place, with message micro-versions; a feature manager plus a global-plugin library whose entries each carry an enable tier; cross-dialogue memory, one-shot temporary chats, and model collaboration; five network-permission tiers, mixed-model routing, a security audit, and browser control through a companion extension.

Fail-closed network recovery, recommended answers, and safe unattended continuation for DeepSeek Harness.

让 DeepSeek Harness 自动重试商汤 429/QUOTA 等瞬时网关故障直到任务完成——启动时给所有 llm-pi-ai provider 热注入加固版 retryPolicy。