Plugins

Compiles LaTeX with the TeX Live installed on the machine: one `latex` tool typesets a formula into an SVG, PNG or PDF and a whole document into a PDF (Chinese text switches to xelatex with ctexart), writes the file into the workspace and returns a Markdown reference to it. A system-prompt section asks the model to write `$...$` and `$$...$$` instead of code fences, and when no TeX Live is found a banner above the composer offers a one-click BasicTeX or TinyTeX install.

Self-narrative memory in five budgeted segments (800 chars, not a full context): a first-person note the model writes to itself, a profile of you aggregated from approved facts, environment facts kept separate from who you are, heat-ranked conclusions, and habits. Every injected fact is traceable to a fact id and can be retired; the profile can be edited by hand, and self-notes stay plain lines you can clear. Facts fade by use and time without ever changing their credibility. Preference-like lines you say are auto-detected as candidates (never auto-approved), with conflicts kept side by side. Recall over past conversations with citations, an incremental index that only decodes newly appended frames, a pointer-required warm layer, a doctor that says which part is broken, a heartbeat proving injection is alive, and export/import so the whole thing moves to another machine.

Long-term memory for dsh with defense-in-depth hardening. Two tiers (project `.dsh-memory/` and user `$DSH_HOME/memory/`), each a plain hand-editable JSON file with bounded entries and tags. Writes are atomic (unique temp file + rename, previous revision kept in a bounded history/) and serialized across processes by a per-tier lock with stale reclaim; a disk-revision check refuses lost updates instead of silently overwriting another writer. Memory text is wrapped in explicit untrusted-data boundaries (forged wrappers stripped on write) to blunt prompt injection from remembered content, load-time validation rejects out-of-bounds stores, and symlink escapes out of the tier directory are refused with MEMORY_STORE_ESCAPE. No content-heuristic filtering and no confirmation step, so headless auto-consolidation stays intact. Install: dsh plugin --profile web add github:RocoFish/deepseek-memory.

Claude Code-compatible persistent memory: the host injects the MEMORY.md index at session start, the model maintains memory files with plain read/write/edit tools, and a Settings page lists saved memories. Same on-disk format as Claude Code; claude mode shares the exact memory directory.

Adds a prompt-refine button to the DSH composer and, when auto mode is on, refines the prompt before the agent starts by rewriting the text with recent session context through a side-channel model call, asks on the native question card when the rewrite cannot be decided, reports each stage as a command card in the chat, and keeps the original text when analysis fails.

Four-tier prompt enhancement for the dsh web GUI: base / standard / professional / system — the system tier always emits a LangGPT-structured Markdown prompt, the iterate workbench rewrites an editable base under an optional requirement, and literal ancho

Security plugin that intercepts file deletion and write commands, requiring user token authorization.

Adds one row to Settings → General with a switch that writes or removes a Windows logon startup entry for the DeepSeek Harness desktop app, and registers a dsh_autostart tool exposing the same status, enable and disable operations. Defaults to a single HKCU Run value, no admin rights; the startup-folder launcher mode adds an optional delay and launch arguments.

Diagnoses why a skill is not picked up — truncated description, broken frontmatter, missing resource files — and checks installed plugins for conflicting loader entries. Keeps a per-profile record so every checkup is diffed against the previous one, groups repeated findings into one root cause, produces rewrite worksheets for oversized descriptions, and lists the plugin patches that touch the skill directory configuration.

Add a WSL directory as a DSH workspace from the sidebar's Add workspace button: pick a distribution, browse the Linux tree, and the directory is registered as a \\wsl.localhost workspace whose sidebar terminals open inside the distribution and whose files the agent can write. Requires Windows 10/11 with WSL installed.

Keep this session working until a wall-clock deadline: one button in the composer picks 22:30, +90m, or tomorrow 07:00, and after every finished turn the plugin posts one continuation message so the session keeps working; nothing is posted once the deadline passes. Stops early on a round cap (400), three consecutive failed turns, or four consecutive rounds that called no tool; a manual stop button is always there. The last part of the watch switches to wrap-up wording, sized to at most half the remaining time so a short watch is not entirely wrap-up. Armings persist to $DSH_HOME/nightwatch.json via an atomic write and are re-adopted after a restart. No slash command is registered by default (enableCommand turns one on); the continuation message carries source.kind cordis-host-runner, not user.

Persistent cost ledger for the DeepSeek Harness: every priced model call lands in a monthly JSONL file; /ledger renders monthly reports, /ledger-export writes CSV, and the agent can query its own spend via ledger_query.

dsh plugin: prefix-cache health guard. Separates prefix rewrites from ordinary context growth using the three disjoint usage buckets, and estimates the tokens re-billed at full price.

DSH plugin: a "No API key required" toggle on saved hand-declared provider cards that writes Authorization: Bearer none so keyless local/self-hosted endpoints stop aborting with "No API key for provider"

MemoryBear long-term memory for DeepSeek Harness: per-turn recall, end-of-turn write, and a memory_search tool

Report generator that follows your own style: pick DSH sessions by time window and slice each one to the turns inside it, optionally merge Git commits from several repositories filtered by author, then a new session reads the evidence with read-only collector scripts and writes a daily, weekly or monthly Markdown report matching a pasted sample; finished reports stay in a local library.

Novel writing engine for DeepSeek Harness: state machine, gates, rulepacks and ledgers moved from prompts into code.

A companion plugin for dsh-pet: gives the desktop pet a persona card, Chinese reasoning lines, click reactions, and a daily diary she writes herself.

Local development facts for Harness: environment and runtime-health probes, git status/diff/log/commit, write-time snapshots with version-by-version restore, and project checks for Node, Python and Go.

Replies to DingTalk direct messages as your own account — reads the bound DSH workspace, writes the answer, and sends it back through the dws CLI.

One MCP tool surface over two Windows desktop-automation backends (dsh-computer-use-win and windows-mcp). Both must be installed first: dsh-computer-use-win as a DSH plugin, windows-mcp via uv tool install windows-mcp --python 3.14, then run the bundled scripts/patch-backend.mjs once (it rewrites the local windows-mcp source; without it screenshot/read_webpage/launch_app fail) and disable the standalone wincu mount. It merges their 44 tools into 29 by dropping 10 duplicates, 4 dangerous ones (PowerShell, Registry, FileSystem, Process) and 1 tool-management entry, and blocks dangerous calls before forwarding through a reversibility-graded guard.

Adds a Continue button to the composer dock and resumes automatically when a turn is cut off by a network drop or an upstream output-limit cut. A resident link monitor asks every 10 seconds whether the network is back, because the link returning fires no event, and a turn that completes normally counts as proof the link works. The click goes through the host submission pipeline — beginSubmission plus session.prompt — so the message carries real user attribution instead of faking one, and a genuine refusal is reported rather than retried. On startup it also scans recent session logs for both fingerprints of a host that left mid-turn: an unclosed turn, and a turn closed as disposed, which is what a normal shutdown writes. A turn the user stopped by hand is never resumed.

Writes one short line into the system prompt only when a stress, energy or reward value crosses a threshold; the values rise on real session events and decay over time.

Token usage dashboard for DeepSeek Harness. A dedicated settings page shows cumulative input / output / cache-read / cache-write token totals, a green GitHub-style daily heatmap colored by usage, and a pie / bar chart toggle for per-model (provider/model) usage shares — all read from local session logs, no upstream API calls.