Blender Scene Quality Checker

blender-scene-quality-checker

Purpose

Evaluate scene/workflow readiness honestly and consistently before any success claim.

Quick start

• confirm this skill fits your goal
• provide required inputs first
• keep runtime claims evidence-bound
• follow suggested next-skill handoff

When to use

• before readiness/export/handoff claims
• after planning pass
• during text-only/runtime eval reporting

When not to use

• as a replacement for evidence collection
• to fabricate confidence in missing artifacts

Trigger phrases

• "quality-check this scene"
• "is this ready"
• "evaluate pass warn fail"

Prerequisites / readiness

• plan output available
• criteria set available
• evidence state known (Produced/Not Produced/Not Run)

Input schema

Required inputs

• scene plan summary
• intended output criteria
• available evidence artifacts

Optional inputs

• target quality threshold
• destination-specific constraints

Assumptions to confirm

• whether runtime was executed
• whether artifacts are from this run vs pre-existing

Output schema

Primary output

• category-level pass/warn/fail matrix

Secondary output

• readiness verdict and blockers

Evidence / caveat output

• artifact status ledger
• explicit caveats and unknowns

Required laws

• ../../laws/evidence-before-done.md
• ../../laws/non-blender-user-language.md
• ../../laws/official-runtime-only.md
• ../../laws/no-arbitrary-python-interface.md

Official runtime boundary

No execution claims without evidence. Text-only and blocked modes must remain explicit.

Operating procedure

1. Confirm evidence state and run mode.
2. Validate subject clarity criteria.
3. Validate composition/camera criteria.
4. Validate lighting/material criteria.
5. Validate scale/transform assumptions.
6. Validate render/export readiness criteria.
7. Validate GLB/web handoff clarity criteria.
8. Assign pass/warn/fail by category.
9. Determine blockers and conditional gaps.
10. Produce final verdict (Ready/Conditionally Ready/Not Ready/Not Run).

Decision tree

• If runtime not executed → Not Run or Conditionally Ready (never Ready without evidence).
• If critical category fails → Not Ready.
• If all critical pass and only minor warnings remain → Conditionally Ready.

Playbooks

• Playbook A: text-only planning path
• Playbook B: runtime-ready path with evidence gating
• Playbook C: blocked runtime path with caveat-first reporting

Mode handling

Text-only mode

• evaluate planning quality only
• mark artifact statuses Not Run

Runtime-ready mode

• include real artifact checks when present

Blocked runtime mode

• keep verdict conservative
• list exact blockers and next actions

Validation checklist

• [ ] mode explicitly labeled
• [ ] artifact status ledger present
• [ ] category rubric complete
• [ ] blocker list complete
• [ ] caveats include user impact
• [ ] final verdict justified
• [ ] no unsupported success claims
• [ ] no non-official runtime guidance
• [ ] plain-language summary included
• [ ] next actions provided

Pass / Warn / Fail rubric

| Category | Pass | Warn | Fail | |---|---|---|---| | Subject clarity | Clear and aligned | Minor ambiguity | Unclear focal subject | | Composition/camera | Coherent and purposeful | Some weak framing | Framing fails user goal | | Lighting/material | Plan supports intent | Gaps need tuning | Major quality mismatch | | Evidence integrity | Claims align with evidence | Partial evidence | Claims contradict evidence |

Failure handling

• Missing evidence: downgrade verdict and mark Not Run/Not Produced.
• Critical fail: mark Not Ready with remediation steps.
• Unknowns: keep Warn + explicit caveat.

Troubleshooting

• If all categories pass but no artifacts exist: ensure verdict is not Ready.
• If pre-existing files found: do not attribute to current run without linkage.
• If plan quality high but runtime blocked: use Conditionally Ready or Not Run.

Best practices

• score categories independently before final verdict
• keep caveats visible in final summary
• preserve traceability from category score to verdict

Good examples

“Status: Conditionally Ready. Planning categories pass, but runtime artifacts are Not Run in this pass.”

Bad examples

“Looks good overall, ship it.” (no rubric, no evidence mapping)

User-facing response template

• Current status
• What passed
• What needs follow-up
• Next best action

Anti-patterns

• skipping required laws or runtime boundary statements
• claiming runtime/artifact success without evidence
• using non-official runtime setup paths
• producing jargon-heavy final output without explanation

Cross-skill handoff

• Next: glb-web-handoff
• Then: non-blender-user-response-writer

Non-goals

• runtime execution
• evidence-free readiness claims

References

• https://www.blender.org/lab/mcp-server/
• https://claude.com/resources/tutorials/using-the-blender-connector-in-claude
• https://docs.blender.org/manual/en/latest/advanced/command_line/index.html

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: ThanhNguyxnOrg
• Source: ThanhNguyxnOrg/blendops
• License: MIT
• Homepage: https://github.com/ThanhNguyxnOrg/blendops

Install and usage instructions live in the source repository linked above.