- Home
- Skills
- Agents & Orchestration
- DSH Orchestrator
Works with the AI Agents you already use
DSH Orchestrator
Orchestrate DSH desktop tools and Blender workflows with automated task triaging and runtime verification.
$4
DSH Orchestrator
Example session with this skill installed
Check the current Blender runtime status and generate a previs scene for the 'living_room' layout in the production/blender-previs/ directory using DSH.
- Read your context and instructions
- Compiled the dsh orchestrator
I have verified the local DSH environment and initiated the Blender previs task. The blender_status check confirmed version 4.2 is available and the isolation directory is writable. I have dispatched this to DSH as an asynchronous session to preserve local context.
DSH Task Summary
| Component | Status | Details |
|---|---|---|
| Runtime Check | PASS | Blender 4.2 detected; available: true |
| Connection | ACTIVE | dsh-as-mcp 0.1.5 connected |
| Task Type | DELEGATED | Complex multi-step modeling |
| Session ID | sess_9482_blender_previs | Asynchronous - polling every 30s |
| Output Path | /production/blender-previs/living_room/ | Writable |
Next steps
- Wait for session
sess_9482_blender_previsto return theCOMPLETEDstatus. - Inspect the generated
.blendfile and representative renders in the output directory. - Run the local validation script to check for mesh naming convention compliance.
Connects securely to your tools. The creator never sees your data.
About this skill
This skill optimizes high-stakes development by orchestrating workloads between your primary IDE agent and the DSH desktop environment. It eliminates unnecessary context switching and token waste by routing tasks based on complexity and tool availability. Instead of blindly delegating every request, it performs pre-flight checks on skill registration, runtime availability, and file system permissions to ensure execution won't fail midway.
The skill provides a rigid decision logic for task triage: direct execution for local scripts, targeted dsh_tool_call for specific desktop MCP functions, or full model delegation for complex, multi-file workflows like Blender 3D previs or batch processing. It includes a specific routing table for 30+ built-in Blender skills and handles the entire lifecycle from initial plugin discovery to asynchronous session management and final artifact verification.
What it does
- Task Triaging categorizes workloads to minimize token usage while maintaining output quality.
- Skill Verification checks runtime status and Blender availability before dispatching tasks.
- MCP Orchestration executes DSH desktop tools directly without full model prompts for simple actions.
- Session Management handles asynchronous task delegation with status polling and ID tracking.
- Validation Engine performs post-execution checks on diffs, renders, and samples to confirm success.
How it works
- Triage Stage evaluates if a task requires local execution, a single DSH tool, or a full model delegation.
- Pre-flight Check runs
blender_statusand path permission audits to confirm environment readiness. - Dispatch Phase invokes
dsh_tool_callfor targeted actions orsession_promptfor complex blocks. - Verification Step audits final artifacts against specific success criteria rather than relying on self-reporting.
Frameworks & tools
Built for DSH-as-MCP 0.1.5 host plugins. Integrates with Blender runtimes, TypeScript for control logic, and standard MCP client protocols for desktop tool interaction.
Why this beats prompting it yourself
Manual delegation often results in "hallucinated connectivity" where the agent thinks it can reach desktop tools but fails at runtime. This skill enforces hard checks on local environments and implements an asynchronous polling logic that prevents the primary agent from hanging during long-running 3D renders or batch operations.
Use cases
- Orchestrating complex Blender previs workflows across local and desktop environments.
- Reducing token costs by routing simple file edits to local tools instead of remote agents.
- Managing long-running batch processing tasks via asynchronous session IDs.
- Automating the installation and recovery of DSH desktop bridge components.
Known limitations
Does not support hardcoded Flash model switching. Requires DSH-as-MCP 0.1.5 or higher. Cannot automatically wake the parent session after a timeout without manual polling.
Requires dsh-as-mcp 0.1.5+. No support for legacy bridge 401. Primary agent cannot be automatically woken by DSH completion; requires polling.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 19 days ago
- Software
- tsx
Node.js
Git