Plugins

Agent governance suite: policy-based tool gating (allow/deny/ask with wildcards and priorities), a structured JSONL audit trail, and per-agent token quotas against the host token meter, with state under $DSH_HOME/gov.

Verification-driven self-evolution loop: failure logs become verified AGENTS.md rules; in-session plugin (evolve_learn / evolve_apply / evolve_touch / evolve_recall), tool-verify gate, rule lifecycle, local recall.

Aborts startup when another live dsh instance holds the same DSH_HOME data directory, preventing the concurrent session-log writes that corrupt transcripts.

S3-compatible backend for the spill seam: oversized tool output goes to object storage (S3, MinIO, R2) with hashed session prefixes and unguessable keys, instead of the agent's local disk.

Complement to official Creator mode: stage, validate, live-mount via ctx.plugin, score, and solidify or roll back in-memory tool-plugin trials.

Minimal-anchored agent preset: request #1 keeps the real bash + str_replace_editor pair with auto-injected context stripped, then promotion unlocks the full Standard catalog and posts a one-shot background-jobs notice against sleep-polling habits.

Close and restart controls for DeepSeek Harness Desktop and Web on Windows, in Settings → General.

Collapsible right-docked web terminal (xterm.js) bridged to a host PTY over WebSocket: bash/zsh switching, resizable panel, and fold/unfold that keeps the session alive.

Evidence-backed Skill self-evolution for DeepSeek Harness: redacted trajectory learning, quarantined proposals, arena validation, canary monitoring, automatic rollback.

RLM mode for dsh: recursive sub-agents via a unified rlm() function — spawn, await, and chain child agents as first-class Python calls. Adds a persistent IPython kernel with tools.* bindings and context-as-variable.

Dependency-topology guard: pre-install conflict prediction and post-install detection of duplicate @deepseek-ai/dsh-* copies, version drift, and vendored core services that cause Symbol-key crashes like "Cannot read properties of undefined (reading 'prepare')".

Ecosystem trend dashboard (行情面板): snapshot the dsh-plugin topic and the awesome list into local history, then report weekly trends — new plugins, star gainers, category heat, awesome coverage — plus a keyword radar for new plugins and star surges.

Verification toolkit for DSH agents: evidence-based claim checking against workspace files with line citations, config validation (JSON/YAML), and read-only URL/npm/GitHub submission-readiness probes.

Native Windows system notification (Notification Center toast) when a task completes: goal completion or agent turn end.

Ambient fullscreen status grid for DeepSeek Harness sessions with token/rate graphs, plus ntfy push and alarm notifications when a turn ends, errors, or waits for approval.

Two-way Feishu (Lark) chat bridge for DeepSeek Harness over the official WebSocket long connection: messages route to the bound workspace's latest session and replies return automatically; bot credentials and workspace routing are configurable from the settings page.

Short Web Audio cues when a session needs your attention (approval, question, or plan review) and when a task finishes; synthesized inside the page in two selectable styles with a volume setting — no popups, no OS notifications, no audio files, no network — and one bundle that covers DSH 0.1.7+ (desktop & web) and 0.1.0-rc.6.

Open the current registered Workspace from the Session Header or Workspace menu in a detected or configured editor, with hot re-detection and a remembered browser preference.

Feishu bridge for DeepSeek Harness (derived from dsh-im-bridge): WebSocket long connection, approval cards, group @-mention-only replies, streaming card updates, /stop command.

Bridges DeepSeek Harness lifecycle status, errors, and approval requests to a locally running OpenPets desktop companion.

Adversarial code review — parallel finders inspect a diff through separate lenses (correctness, lifecycle, contract, security), then each finding goes to independent verifiers tasked with refuting it, and one refutation drops it.

Injects allowlisted GitHub Actions and GitLab CI run metadata into agent context without reading logs, credentials, or provider APIs.

GitHub workbench MCP server (dsh bundle): 23 MCP tools for repos, issues, PRs, code review, and search, with zero runtime dependencies and PAT/OAuth auth.

Adversarial multi-lens code review: parallel correctness/security/maintainability lenses attack a diff, with deterministic severity grading, suppression rules and incremental review history.