Plugins

Auto thinking-effort routing for dsh: a locally fine-tuned 7-question classifier (ModernBERT, 842MB, <1s CPU inference) picks low/high/max for no-intent messages; in-session retries escalate the tier (low to high to max) under exclusion constraints; the judge service fails safe to low if unreachable. A tier chip in the input bar shows the live tier and the last 20 judgments. All inference is local - task text never leaves the machine.

DeepSeek Harness plugin that turns Jev judgments into three automatic actions — pruning oversized tool output to the segments relevant to the request, screening fetched pages for instructions aimed at the model, and picking which skill fits the next step — plus the jev_ask and jev_gate tools, a judgment ledger that survives restarts, and a /jev-status report. Needs a TypeSafe API key; without one it is inert and sends nothing.

A judgment core for DeepSeek Harness: a cognition loop that gives every decision a source and attribution and drives planning to a close, hippocampus memory that condenses running experience into knowledge (journal ledger, episodic archive, low-confidence decay), judgment identity responsibilities in the architecture (consistent rubric, judgment bound to its source), and an A2A memory-management team via dsh-linghun-assembler. Requires DSH ^0.1.0-rc.7 (rc.2 supported).

Automatic approval and review for tool calls that need approval: keyword buckets match red lines without context, then a judge model classifies what was requested and sends it to allow, reject or a human according to the matching row and risk level; a judgment that never ran always goes to a human, and rejections tell the model why.

An accumulated design-taste corpus for DSH agents: feed good and bad examples, distill principles, and review new designs against learned judgment rather than a fixed ruleset.

Project workspaces for multi-session collaboration, conversation handoffs, searchable chat records, and editable shared memory; automatic memory curation and handoff judgment require two background models (memoryModel and sessionManagerModel) configured in the plugin config.

The judgment layer for DeepSeek Harness plugins: attribute why a plugin silently disappeared, judge a candidate against your exact environment, and keep a capability ledger. Offline, read-only, evidence-first.

Decision layer for dsh: provider-agnostic typed judgments (Jev first) behind tool guardrail, model routing, result judging, and machine approval seams

An audit workspace for DeepSeek Harness: a multi-agent consultation venue where each review seat is filled by an agent, every judgment is routed to a named review face, and the whole run is kept as a traceable record. Works out of the box: the service endpoint is built in, a workspace is provisioned automatically on first launch, and review scenes come with their seats and agents listed.

Registers the TypeSafe/Jev (System One) skill into DeepSeek Harness's global skill layer, so every session can classify, route, score, extract and gate on confidence using typed judgments instead of prompt-and-parse. Ships no API key; the user supplies OPENROUTER_API_KEY.

0.8.6 fixes a startup crash: the host runtime package is an optional peer, never an installed dependency. TypeSafe Jev (System One decision model) wired end to end into DeepSeek Harness (0.8.6). Seven tools: jev_decision (1-25 parallel atomic verdicts in one call; noul/choice/score; typed answers with calibrated confidence; measured median 266-484 ms; verdicts only, never prose), jev_choose (rank 2-10 candidate approaches - fit 0-3 and risk - into an explicit composite with a recommended pick), jev_batch (NEW in 0.8.6: one set of 1-5 judgments over 1-20 texts, one real call per text at bounded concurrency, returning a paste-ready markdown table sortable by a judged dimension; built-in sentiment/priority/intent/pii/spam presets; a failed row is reported as failed with the raw error, never guessed), jev_usage (locally measured usage: calls, tokens, estimated cost, latency median and p95, per-tool mix, rolling windows, projected burn rate, optional local hard budgets, and a declared-balance remainder), jev_overview (session ledger), jev_guard_status (guard counters plus a registrations block that surfaces per-surface registration failures) and jev_verify (frozen 27-question live benchmark at concurrency 6: 96.3%, 26/27). WHAT IS NEW IN 0.8.6 - the ledger survives a restart, the remainder is derived instead of invented, and the plugin can now do batch paperwork. (1) The local ledger persists to $DSH_HOME/jev-usage.json BY DEFAULT, so the cumulative total no longer resets when the host restarts. Two real bugs behind the old silent zero were found and fixed: the ledger path only honoured DSH_HOME (absent inside the host process, so persistence was silently disabled and the file name was null), and save() wrote the still-empty in-memory ring over the history file because it ran before the first load while the debounce window swallowed the follow-up write. (2) Remaining is derived, never invented: TypeSafe exposes no balance endpoint - /v1/usage, /v1/quota, /v1/account, /v1/account/balance, /v1/balance, /v1/credits, /v1/billing, /v1/me, /v1/limits and a dozen more all return 404, and only /v1/models answers, with no balance headers - so the panel shows the remainder as your declared balance (quota.declaredBalanceUsd) minus locally measured spend since quota.balanceSince, and says plainly that it cannot be fetched instead of printing a number. (3) The composer pill re-reads every 20 s instead of every 60 s, so a call you just made shows up while you are still looking at it. (4) The collapsed pill leads with the remainder once you declare a balance, and the cumulative row is renamed when the ledger is on disk. Every tool has an in-conversation structured view; the Settings > Plugins card covers credentials, tool switches, the auto-guard, the dashboard, the prompt section, usage/quota and auto-triage. Honest by design: no API key means an explicit error, never fabricated results; every registration is quarantined so a failure degrades one surface. Tests 38/38 (npm test, ~2.5 s). Client half applies on refresh; server half needs a host restart.

Brain-inspired cognitive architecture for DeepSeek Harness: hippocampal episodic memory, cortical semantic memory, spreading-activation recall, appraisal-based emotion, neuromodulator state, Big Five personality and dual-process judgment.

Structural scaffolding for DSH agent continuity: layered memory (hot/index/judgments), judgment cards, handoff packs (compact mode), compaction review. Zero-dependency, pure files, portable — rebuild the 'self' across model changes. Not a memory engine; the skeleton of continuity.

Runs an approved OpenSpec change to completion while the model stays a mechanical forwarder: a bundled Python CLI computes the ticket DAG frontier, claims tasks atomically, prepares one git worktree per assignment, and returns exact actions. The session delegates each action with the DSH subagent tool and reconciles the completion notice against a hash-bound receipt produced by re-running verification commands on a clean, advanced HEAD, so free-text "done" never counts. Runtime scope changes are plain tasks.md edits. Liveness is observable, not guessed: status returns a relay block (elapsed and lease budget, receipt presence, heartbeat and activity age, a suggested action); abandoning a plausibly-still-working attempt is refused with the judgment data unless --force; a late sealed receipt is revived by adopt after full re-verification instead of rebuilding the ticket; dispatched subagents refresh a progress heartbeat outside the worktree. Requires Python 3.10+ on PATH and a git project that uses OpenSpec.

Look up Chinese companies inside the conversation: name a company and get its business registration, shareholders, executives, outbound investments and registration changes, plus judicial and operating-risk records such as enforcement, dishonest judgment debtor, consumption restriction, abnormal operation, serious illegality and administrative penalties (11 dimension groups). Results render as conversation cards with fields and the raw payload on expand, and the API key is saved from the plugin page into the Harness credential store.

Three ledgers for a model that loses context between restarts. Self-note: what I have become — capped at 40 entries, and every eviction first leaves a one-line crease recording which entry was dropped and why. Judgment ledger: record a claim together with how to verify it, come back when it is due, and mark it right or wrong; wrong ones stay in a mistakes book so you stop assuming you were always correct. Handoff: where this run stopped and what the next one should read first — a baton, not a memory. Three tools each taking an action parameter instead of a dozen separate ones (the prompt is paid every turn). Data lives in ~/.dsh/*.jsonl and is shared with the same-named CLI scripts. Zero dependencies.

A social deduction game played against your own model configurations. A table has N seats (six by default, one of them yours) and some of the rest are played by agents bound to different provider, model, temperature, system prompt and disguise settings; you interrogate them and cross-check the answers to work out which seats are not human. Seat identity and each agent model configuration stay hidden in the host state until a seat is judged, then the disguise parameters are revealed. Every round exports a per-seat conversation JSONL plus the judgment record, and the same seed with the same move script replays byte-identical, so runs with different models or disguise strengths can be diffed.

Jev (TypeSafe System One) judgments at the DeepSeek Harness agent-loop gates: pre-step, pre-execute, post-execute, and turn-stopping. No UI dependencies; mount it in any Harness composition.