Plugins

Red-team engagement mode: send one target organization name; a planner session runs five execution roles (recon, asset triage, vulnerability discovery, exploitation, internal pivot) with configurable concurrency; preflight checks skills/resources; scoring uses four stages (recon, internet, boundary, internal) with server-side caps and privilege-wins dedup; findings stay in a local SQLite fact base. The persistent right-side console has nine pages and includes asset mapping, agents, sessions/tunnels, findings, score targets, reports, POC knowledge base and skill library. v0.15.2 adds automatic HTTP evidence/action records, batch ingest, scope check, tunnel pivot probe, data proof, cleanup ledger and T0-T3 asset admission to prevent scope drift. 59 redteam_* tools; 38 bundled skills.

Academic writing guard (EN/ZH bilingual): removes AI-style defensive writing, protects scientific evidence (numbers, citations, claim strength, null findings, scope) during polishing, and calibrates writing toward a target journal (Journal Profile + Journal Fit). Deterministic local regex/statistics, zero network/LLM; writing_rules / writing_audit / writing_style_profile / writing_journal_profile with auto-audit on paper writes.

Installs four multi-agent math-research presets for DSH: v2 (probability-driven pipeline with multi-verifier debate), v3 (paper-style Markdown knowledge base with a planner agent and a reusable method library), v4 (persistent self-organizing residents that message and meet), and v5 (a research institute with an academician who decomposes and assigns work, voting researchers, temp workers, group chat and a compare-and-set task board); all four support checkpoint resume, human intervention, and an optional Lean formal-verification switch (off/encourage/require) whose passing proof turns the vote into a fidelity check of the Lean statements.

Novel-writing guardrails: 20 novel_* tools enforce fact-ledger consistency, phase and chapter gates, machine audit with deterministic metrics, and proposal-only revisions; a right-panel workbench covers reading, TTS playback, polish, proofread, continuity checks and batch drafting, and a bypass engine runs internal steps off the main conversation.

A web-novel writing engine for DeepSeek Harness: parallel drafting, outlining and brainstorming with a separate DeepSeek key, lore and foreshadowing management, semantic vector retrieval, a corpus library, usage ledger, mechanical proofreading, and a local serialization plan.

Live security beacon for DSH that follows tool activity. Click to view call statistics, risk provenance chains and blocked actions, with read-only task restrictions and tool metadata change detection.

Math Agent Framework for DeepSeek Harness: symbolic derivation engine (CES / quadratic forms / ODE / PDE), SymPy verification, Lean 4 compiler-verified formal proofs, and QED-style multi-agent adversarial verification, exposed as native DSH tools through a TS shell bridging a Python engine. Requires cloning the math-agent-framework repo and `pip install -r requirements.txt`, with MATH_AGENT_HOME pointing at the repo root. Every result carries provenance (engine versions, seed, tolerances) and is marked untrusted data.

Multilingual writing, chunked long-form collaboration and continuity review, with on-demand DSH delegation discovery. Requires Python 3.9+ on PATH.

A `/backstory` command and tool that annotate each line with the git commit that last touched it and the agent turn + prompt that wrote it, from a persistent per-line ledger (drift-proof by content hash), DSH-* commit trailers, or the live session log.

Real-compute verification gates: math simplification, logic proofs, regex validation, E-prover FOL, state-machine checks, and code repair, backed by Go-compiled binaries with prebuilt Windows executables.

Pins the first prompt you sent in the current session in a scroll-proof bar above the conversation, with one-click copy, a collapsible panel listing every prompt you sent, and eight switchable looks you can pick from inside the panel.

Provenance-aware execution security for DeepSeek Harness. Track sensitive data across tool calls and stop risky side effects before execution.

Model-facing image_describe (识图) tool over the DashScope OpenAI-compatible API (qwen3.7-flash), plus a paste bridge: on text-only sessions, pasted images auto-convert to file paths at send time and render back in the transcript, so they never trip image admission. Bring your own DASHSCOPE_API_KEY; endpoint/model/budgets configurable, redirect-proof HTTP client, works in every agent preset.

Independent read-only acceptance layer: spawns a read-only verifier before each top-level turn closes and steers non-pass gaps back into the agent.

Auditable mathematical proving for DeepSeek Harness: a local stdio MCP server with 23 machine-checking tools, a bundled workflow skill, and the /prove and /audit-proof commands. proven, refuted and inconclusive are never conflated, and unproven steps are reported instead of hidden. Needs Python 3.10+ with pip install -r requirements.txt on first run; the 23 tools are unavailable until then.

Session-scoped research evidence ledger with sources, linked claims, open questions, and explicit user review; provides a tool, slash command, and conversation panel, with selected-claim Markdown export that omits excerpts, review notes, and local paths by default; does not fetch sources or verify their truth.

Session-scoped delivery receipts for local artifacts: compare current file hashes with a recorded baseline, keep agent claims separate from explicit user acceptance, flag changed or unavailable files, and export a Markdown receipt through a tool, slash command, and conversation panel.

Prove that a local image edit left every pixel outside its mask untouched. Original, generated and mask PNGs go in; a greppable `MASK_PIXEL_LOCK=PASS|FAIL|EMPTY` verdict, a difference heatmap, per-channel statistics, a bounding box and the largest changed pixels come out, plus a self-contained HTML report and `report.json` with a SHA-256 file hash, a canonical protected-pixel digest and a proof digest binding both to the policy. Supports edit and lock mask polarities, a mask threshold, tolerance and alpha-ignoring; PNG color types 0/2/3/4/6, bit depths 1-16, all five filters and Adam7. Zero dependencies, local only — no model call, no network, no upload.

Turn English audiobooks (MP3) and documents into listening lessons: transcribe every sentence locally with the speech model DSH voice input already downloaded (about 241 MB, fetched on first use if missing; ffmpeg required), proofread and explain each sentence with your DSH model, then drill with blind listening, intensive listening and spaced repetition, with Chinese-dubbing export and offline Kokoro voices; ships a sidebar panel and four AI tools.

Session-scoped research evidence ledger with sources, linked claims, open questions, and explicit user review; provides a tool, slash command, and conversation panel, with selected-claim Markdown export that omits excerpts, review notes, and local paths by default; does not fetch sources or verify their truth.

Session-scoped delivery receipts for local artifacts: compare current file hashes with a recorded baseline, keep agent claims separate from explicit user acceptance, flag changed or unavailable files, and export a Markdown receipt through a tool, slash command, and conversation panel.

DSH plugin: cross-session call center for agents. session_message delivers messages between DSH sessions (idle: visible full-text wake; busy: mid-turn-safe notice; plus runtime-context notes) and resolves targets by registered contact name (stale-id-proof

Multi-instance ComfyUI orchestration: probes every configured ComfyUI endpoint for its GPU, free VRAM and queue depth, then dispatches a generation job to the idlest eligible instance. Ships three tools (comfyui_farm_status, comfyui_farm_pick, comfyui_farm_run). An instance is eligible only if it is reachable, has at least the requested free VRAM, and its queue is no deeper than the configured ceiling; eligible instances are ranked by free VRAM minus queue depth times a weight, so an idle shallow queue wins over a busy large one. Zero runtime dependencies. Optional retry on probe failure, because a single timeout across an SSH tunnel is not proof that an instance is down.

Verifiable authorization, execution evidence, and offline delivery verification for DeepSeek Harness