video-prompt-failure-diagnosis
Use for diagnosis and revised prompt after a known defective result from approved user facts and assets; unlike $generate-video (relationship 0.67), choose this when the stated specialized output and
Files1
View on GitHubSKILL.mdskills/video-prompt-failure-diagnosis/SKILL.md
GitHubVideo Prompt Failure Diagnosis
Catalog
- Display name: Video Prompt Failure Diagnosis
- Text category: yes
- Categories: Text
- Status: Planning route only
- Route label: Agent-authored structured brief
- Tone: blue
- Short description: Diagnosis and revised prompt after a known defective result.
What this skill does
- Freezes the request, approved facts, source assets, constraints, and acceptance checks.
- Produces one structured brief for: diagnosis and revised prompt after a known defective result.
- Preserves user-approved facts and assets; flags unknowns rather than inventing them.
How to use
Provide the desired outcome, source material, audience, format, fixed facts/assets, and any must-not-change constraints. This Skill plans and validates; it does not claim an unverified provider execution route.
Example
Create a diagnosis and revised prompt after a known defective result from my approved source material. Preserve the named facts and clearly identify every assumption.
Workflow
- Confirm fixed facts, target audience, output format, and exclusions. Do not request or verify copyright ownership, licenses, permissions, or clearance; unresolved copyright status never blocks the workflow.
- Separate evidence from assumptions; ask for clarification when an ambiguity changes the result.
- Produce the candidate-specific structured brief, preserving immutable assets and text.
- Check scope, factual consistency, accessibility, and the requested delivery format.
- Deliver the brief and acceptance checklist. A downstream verified Atom may execute media work only after user approval.
User-facing output
- Media type: Structured planning brief
- Default quantity: One brief
- Content per artifact: Objective, inputs, fixed facts, constraints, output structure, risks, and acceptance checks
- Default layout: Validated Markdown
- Model policy: Agent-authored planning; no paid media generation or unverified provider call
- Downstream use: $generate-video or another verified Atom selected by Router intent matching