Plugins

Registers one agent tool, plugin_guard, that statically scans a DSH plugin package before it is installed and reports a verdict with the reasons: a missing or broken dsh.bundle.patch, a missing dsh.permissions block, install scripts, dangerous APIs and obfuscation, dependency risk, and source that is not indexed and therefore cannot be verified. Read-only: the scanned code is never imported, installed, evaluated or run.

Checks installed DSH plugins for version-declaration and API risks against the current host, with source locations when available, certainty, impact basis and coverage limits. Adds a compatibility settings panel, plugin-detail findings/badges, a compat_check agent tool and a skill provider.

Config snapshot and rollback for DSH: auto-captures before risky changes, ships a standalone restore script that works even when DSH won't boot. Windows only; the restore script is a .bat and needs Node.js.

Evidence-first reuse skill for DeepSeek Harness: before implementing, porting, choosing dependencies, or troubleshooting a tool, it searches the local project and machine for existing solutions in layers, judges reuse on checkable evidence, refuses reuse when essential requirements or risks are unmet, and reports what could not be verified. Ships as a DSH bundle that registers the skill on ctx.skills.

Reads and writes WPS 365 cloud documents (smart documents and .docx/.xlsx/.pptx) through the locally installed wps365-cli, with conflict checks before writing; requires a WPS 365 enterprise account.

Release gate for AI answer quality. Two channels: replay an existing VeriScope archive through the original Python quality gate, or run a threshold gate over a Ragas per-row export with Node only — no VeriScope, Python or pandas. Returns PASS, FAIL or REVIEW with source evidence and a re-checkable receipt; supports per-row thresholds and exact decimal comparison. It never re-evaluates and never calls a model, and REVIEW never becomes PASS.

Cache-hit metrics, request-prefix risk confirmation, and per-step cache-miss diagnostics for DSH.

Watches a turn's thinking and output streams and cancels the turn when the model starts repeating itself. Four detectors: a verbatim cycle, a counting loop whose lines differ only by a number, repeated lines separated by other text, and a user-supplied phrase list. A catch raises a manual-dismiss alarm card in the global overlay with an optional tone, and a Monitor tab holds the switches, thresholds and hit history. Detection is local string comparison with no model calls, so it costs no tokens.

Writes a model-generated annotation into every DSH escalation approval prompt — what the command does, what the model intends, how risky it is and why, and what each choice means — and keeps a searchable history of past approvals at /dsh-approval/history, with a diagnostics page at /dsh-approval/diag.

Conflict diagnosis and repair for DeepSeek Harness plugins. Builds a capability graph of the running Loader profile and of the on-disk cordis.patch.yml layers, comparing imports, peer ranges and config against the host that is ACTUALLY running rather than the profile's possibly-stale copies, and reports the conflicts that fail silently: a fiber stuck PENDING on a service no plugin provides, a browser half injecting a client module nothing provides, a tool id two plugins both register, duplicate Loader entry ids, a row disabled by one layer and re-inserted by another, an import of a host subpath the running host does not export, and a cycle among inject declarations. Every finding carries its evidence, a confidence grade for who owns a name, and a repair with a risk grade — verify it with a dry-run difference table, or take the individual .patch.yml files that --fix-out writes (never auto-loaded, so picking wrong breaks nothing).

Host-side long-task detection for small-context local models: long outputs are scaffolded into a parts skeleton (one part per reply, then assemble), and multi-step work into a PLAN.md ledger with notes/ (plan first, one step per reply). Detection is heuristic and host-side, so it does not depend on the model judging its own limits.

Pixel-level screenshot diff for visual regression. Two local PNGs in; a machine-readable first line (`PIXEL_CHANGED=true|false`, greppable as an assertion), then the difference ratio, per-region bounding boxes, maximum channel delta and mean absolute error, plus a self-contained HTML comparison page. An ignore-region list with absolute or fractional coordinates excludes clocks, cursors and other per-render noise. It deliberately does no content recognition: it answers only "did this change", which is the part that can be made deterministic. Entirely local — no model call, no network, no upload.

E-commerce plugin for DeepSeek Harness — product search, comparison and checkout across Taobao/JD/Pinduoduo, via the yuantoutao Streamable HTTP MCP server.

Connects Trae enterprise models to DeepSeek Harness from a personal access token alone, calling llm_raw_chat directly with no trae2api or traecli process; fetches the model catalog from upstream on startup and shows remaining quota for personal packs and enterprise seats.

ChatGPT-like conversation layout for the DeepSeek Harness Web client: a 768px centered column, right-aligned neutral bubbles for user turns, full-width assistant text at 16/28, and rounded code blocks with their own copy banner. Adds semantic inline markup (`==confirmed==`, `!!risk!!`, `++key word++`, with bare 4xx/5xx codes highlighted automatically) and line-start check/warn/tip callouts; a fast/thinking chip in the composer that switches the model reasoning effort through ctx.modelDirectories; and sidebar contrast, where the selected session gets a tinted row with an accent bar and workspace folder icons are colored.

Visualizes prompt-cache hit rates and model request activity in DSH Web, with request details, comparison, and conversation navigation.

安全隐患台账→《工贸企业重大事故隐患判定标准》（应急管理部令第10号）条款映射提示（仅提示差异，不作出定性结论）

Checks an FMEA worksheet for element completeness and for the arithmetic that follows from it - severity, occurrence and detection against the resulting risk priority number - following published FMEA method standards. Reports inconsistencies; it does not judge whether the ratings themselves are appropriate.

Checks an electrical work permit for the elements it must carry and for the issue and permission sequence the current electrical safety regulations require. Reports the differences it finds; it does not judge whether the work was safe or whether the permit was validly authorised.

Flags ICD coding pairs that need a coder's attention - dagger and asterisk combinations, combination codes - against GB/T 14396-2016 and published ICD usage notes. Reports the pairs it finds rather than resolving them; the coding decision stays with the coder.

Blocks AI coding agents from editing files without evidence and from running unapproved high-risk commands, and tells quoted dangerous strings apart from commands actually being executed.

A pale rose-and-gold Mushoku Tensei skin for the DSH web GUI: an arcane sigil and two full-body Eris figures behind the conversation, a young Eris standee and gilt plugin glyphs in the sidebar, and its own settings card.

Read, audit and diff Protocol Buffer definitions without installing protoc: a zero-dependency .proto parser exposing six tools (status, outline, describe, diff, audit, next_number). diff classifies every change as breaking, risky or additive and names the wire-format consequence; audit catches what protoc accepts but a reviewer should not — duplicated or illegal field numbers, reserved-range collisions, unresolved types, and unreserved gaps below the high-water mark.

Rewrites the draft in the DSH composer into a clearer, more executable prompt through the harness's own LLM routes, showing an original/optimized comparison before you adopt it.