Plugins

White-box AGI architecture exploration: metacognition (self-cognition loop), continual learning (knowledge flywheel), world model (condition space, spatiotemporal memory graph), self-improvement (bootstrap discipline), zero-LLM white-box pipeline, and auditable trust guardrails.

Local-only project memory with zero bundled runtime dependencies or external services: JSON as the source of truth plus generated PROJECT_MEMORY.md, token-bounded session-frozen Hot Memory for prompt-prefix cache stability, cached cross-project BM25 recall, trusted session/turn provenance, explainable pre-write duplicate/conflict governance, editable importance/pin/tags/supersede/archive lifecycle, configurable top-level turn-end distillation, and a bilingual Web GUI for search, inspection, diagnostics, and live settings.

Local-first learning mode: cross-session learning threads with per-source explanations.

Learning mode that explains each agent step (WIP).

Select any text in a conversation and tap the Explain button to get a streaming Markdown explanation bubble, with recursive follow-up questions.

Compact Codex coding mode for DSH: light prompt guardrails against GPT over-explaining, with an editable prompt and native Codex-style tools.

DSH Desktop enhancement plugin: it puts verifiable facts on the table. Constraints are injected only when relevant, so idle chat costs no extra tokens; injection is layered (session-stable text in the system prompt, volatile content as a tail snapshot) and the prompt cache stays valid. Discipline: each turn is classified (question / research / discussion / diagnosis / execution) with boundaries enforced - no file edits on question turns, no unsolicited fixes on diagnosis turns; citing code lines not opened this session, negative claims about unseen symbols, requirements that do not match the deliverable, and dressing up “I did not check” as “waiting for your decision” are all called out; replies are written for humans - a code, field name or id must be explained in plain words the first time it appears. Memory and knowledge: 75%/90% context warnings plus an exported session recap (say "continue" in a new session); project knowledge accumulated across sessions by working directory, auto-captured and deletable by id. Inspectable artifacts: task contract / change record (auto-logged, unverified items flagged at delivery) / hypothesis record (conclusions must pass an evidence gate) / design decisions. Dashboard and personas: routing, trigger hits and block assembly are written to lume-metrics.jsonl and queryable via lume_metrics; personas are distilled from chat logs, novels, scripts or character sheets, with long-term and temporary memory, a memory star map and persona card export/import.

Shows a floating toolbar above text selected in a conversation, with copy, quote, ask, explain, translate, summarize, and custom prompt actions.

Feynman learning-mode plugin: teach → teach-back → judge → re-explain loop rendered as rich HTML lesson cards (mermaid diagrams + shiki code highlighting).

JSON Schema validation: validate/paths/explain/normalize.

Test/extract/safe-replace/statically explain regexes without executing code.

Explain failed DeepSeek Harness agent runs locally, then produce a redacted recovery plan. No model required.

Source-aware prompt injection guard for DSH: tracks untrusted turn context, detects injection signals, scores sensitive sinks, and blocks high-risk tool calls with explainable audit logs.

Bounded local memory with CBDC authority gating: SQLite + FTS5 claims, scoped recall with explainable use/verify/ignore decisions and a full audit trail, /memory commands, ≤3-claim/1200-char injection per call, no extra model call.

Explain a selected word or sentence from a DSH conversation in a floating window: a channel of its own that leaves the session untouched, and the answer can be pinned to a session of its own.

Adds an Explain button to the approval card: one model call returns exactly three lines reading the pending operation, covering what it does, which files or system state it reads, writes, modifies or executes, and a risky/attention/safe verdict with a one-sentence justification. It also renders the detail area for `write` and `edit` calls, which the shipped renderer leaves blank because it only reads a `command` field.

Go (weiqi) review coach: parses SGF records, backfills winrate and score analysis with a bundled KataGo engine, explains the costly moves in the chat, and writes the commentary back into an SGF copy; a web panel shows the board with problem moves, AI choices and variations.

A tiny octopus desktop pet that explains AI-jargon in plain language for the DeepSeek Harness web GUI.

Update copilot for DeepSeek Harness: one scan covers the DeepSeek Harness core and every profile plugin; explains what changed and how risky each update is, then updates only what you confirm.

Live per-task token cost widget: peak/valley pricing windows, projected post-hike costs, sortable per-task cost table, multi-currency (COP/USD/CNY + ~40 more) with FX rates, and one-click streamed LLM explain of expensive tasks.

Select text in the DSH web UI to stream a translation and an explanation of what it means in that specific context, read against the conversation it came from; the panel keeps a replayable history and can be promoted into a full session.

Generates 3Blue1Brown-style Manim animations to explain answers, with a right-column gallery and inline multi-modal file publishing.

OpenReelbench, the open video creative bench for DeepSeek Harness: a gated four-stage pipeline (brief -> script -> assets -> compose) that turns one sentence into a narrated explainer video; generation via ComfyUI workflows, assembly with FFmpeg. / DSH 开源

Select text in any Windows app to explain or translate it with the Harness agent, answered in a native always-on-top overlay.