dsh-ark9canvas
Image generation workbench and agent tool for DSH: one tool (ark9_generate_image) paints text-to-image and image-to-image via any OpenAI-compatible API. Channels are entirely user-configured - no bundled endpoint, you bring your own baseURL, key and model name (manual input supported). Agent-initiated generations are approval-gated by default: requests queue in the floating panel and nothing bills until you approve; approval is non-blocking (the agent replies "awaiting your approval" immediately, and results, denials or timeouts are injected back so the agent reports them in the next turn). Floating glass workbench with five tabs (Generate / Approvals / Prompts / History / About), aspect-ratio grid with quality-budget + 16px-alignment sizing, transparent background, batch up to 10 images as aggregated sub-tasks, prompt library with custom JSON sources fetched through a host proxy, persistent generation logs with one-click retry, multi-channel aggregation with per-channel model fetching, bilingual UI (zh/en), stroke-SVG icon set, config import/export. Registers as a Better Sidebar tab when dsh-better-sidebar is installed; stacks above the dsh-cua FAB when both are present.
On this page
Review info: no independent structural review record is currently matched.
Community signals
Category
Tools & Capabilities
GitHub stars
1
npm downloads
1,874
Discovery registry added
2026-09-03
Review information
Scanner status
No review data
Review mode
No review data
Last checked
No review data
Build approval required
No review data
Lifecycle scripts
No review data
Runtime artifacts committed
No review data
Documented profiles
No review data
Why this site does not provide the install command
A third-party DSH plugin is executable local code. Even when structural review evidence exists, registry data, stars and metadata cannot prove the code is non-malicious. Read the current author repository and perform your own review in a disposable environment.
Install and usage source
This site links to the author's source, npm page and community registry instead of redistributing commands. A plugin may add Web UI, Settings, agent tools or background capabilities; follow the author README for the actual entry point.
Minimum checks before install
- Confirm package.json, the bundle patch and README belong to the same author repository.
- Inspect preinstall/install/postinstall/prepare, child_process, shell, filesystem, network calls and credential access.
- First install into a disposable profile/workspace without production credentials or an important repository.
- If pnpm asks for allowBuilds/build approval, treat it as permission to execute third-party code during installation; do not approve mechanically.
- For reproducibility, pin an exact npm version or Git commit yourself instead of depending on main/latest.
“Third-party” only means the plugin is not an official DeepSeek plugin. Structural checks improve transparency but cannot prove business logic is non-malicious or fully audit dependencies, future updates or runtime behavior.