Prompt Preflight Qa
Run a credit-saving QA pass on storyboard or video-generation prompts before you spend model spend on weak hooks or missing constraints.
Overview
Prompt Preflight QA is an agent skill most often used in Validate (also Build content pipelines, Grow creative ops) that audits video and storyboard prompts for controllability before generation.
Install
npx skills add https://github.com/postplusai/postplus-skills --skill prompt-preflight-qaWhat is this skill?
- Classifies prompt type: storyboard grid, full video request, no-reference, or reference-led replication
- Scores controllability: opening strength, first-three-second hook, drift risk, product timing, and negative constraints
- Checks reference contract, explicit bindings, UGC realism, and long-video segment plans
- Outputs a concise readiness report (optionally machine-readable) and routes blockers upstream—not full rewrites by defau
Adoption & trust: 1 installs on skills.sh; 8 GitHub stars; 3/3 security scanners passed (skills.sh audits); trending (+100% hot-view momentum).
What problem does it solve?
You have a storyboard or video prompt draft but cannot tell if it will burn credits on a weak first frame, drift, or bad product timing.
Who is it for?
Solo builders running PostPlus-style generative ad workflows who want a structured prompt gate before image or video model calls.
Skip if: Blank-slate concepting with no prompt draft, or judging finished creative quality after generation has already run.
When should I use this skill?
A storyboard or video-request prompt draft exists and you need to catch weak first frames, drift risk, missing constraints, or bad product timing before spending model credits.
What do I get? / Deliverables
You receive a concise preflight readiness report with classified blockers so you fix upstream before invoking generation skills.
- Concise preflight readiness report
- Optional machine-readable preflight artifact with blocker list
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Preflight review belongs in Validate when you have a draft prompt and need go/no-go signal before expensive generation—not post-render critique. Prototype subphase covers pre-production creative specs and prompt contracts for UGC-style ads and long-form video segments.
Where it fits
You freeze a storyboard grid prompt and run preflight to confirm the first-three-second hook is legible before any image gen.
You wire PostPlus routing and need a readiness report on reference-led replication bindings before the video API call.
You batch UGC ad variants and preflight each full video request for product timing and negative constraints to avoid wasted spend.
How it compares
Use as a pre-generation checker focused on constraints and hooks, not as a post-render creative director or full prompt author.
Common Questions / FAQ
Who is prompt-preflight-qa for?
It is for builders who already drafted storyboard or video prompts and need a controllability-focused QA pass before paying for generation.
When should I use prompt-preflight-qa?
Use it in Validate before first generation; also when refining Build-stage creative pipelines or Grow-stage ad batches whenever a prompt draft exists and credit risk is high.
Is prompt-preflight-qa safe to install?
It is read/analyze oriented on prompt text—review the Security Audits panel on this Prism page if the skill bundle touches external APIs in your stack.
SKILL.md
READMESKILL.md - Prompt Preflight Qa
# Prompt Preflight QA ## Use When - A storyboard prompt or full video request already exists. - You need to catch weak first frames, missing constraints, drift risk, bad product timing, or generic ad-like language before spending credits. - The output should be a concise readiness report, not creative approval. ## Do Not Use When - Do not use this after generation to judge finished creative quality. - Do not rewrite the whole prompt unless the user asks; identify blockers and hand off to the right upstream skill. ## Core Rule Judge prompt quality by controllability, not by impressive prose. ## Default Workflow 1. Classify the prompt object: storyboard grid prompt, full video request, no-reference draft, or reference-led replication draft. 2. Check opening strength, first-three-second hook legibility, viewer question, visible evidence, promise delivery, product timing, negative constraints, reference contract, explicit reference bindings, UGC realism, output format, and long-video segment plan. 3. Create a machine-readable preflight report when it is useful. 4. Return a concise verdict with risks, missing fields, likely drift, fixes, and whether the prompt can run now. ## Blame Rule Point the weakness to the correct upstream stage: - bad route -> `media-router` or `video-request-architect` - missing or misfit hook logic -> `video-request-architect` - weak hook decode -> `reference-decode` - weak panel, beat, or request logic -> `video-request-architect` - unclear learn/do-not-copy boundaries -> `reference-contract-builder` ## Output Shape The artifact contains `verdict`, `canRunNow`, `missingFields`, `majorRisks`, `likelyDrift`, and `fixNow`. ## Stop Conditions - Stop when required user intent, source evidence, or owned input artifacts are missing and guessing would change the result. - If an owned CLI or script command fails, report the exact error and stop. Do not bypass the failure with metadata-only answers, readiness probing, local payload rewrites, fallback providers, or unpublished tools. ## Handoff - Return `report.json` or the stdout JSON plus the smallest upstream fix. - After fixes, rerun this preflight before generation. ## Public Command Boundary - Choose the smallest matching command or workflow from the user input and run it directly. - If an owned CLI or script command fails, report the exact error and stop. Do not bypass the failure with metadata-only answers, readiness probing, local payload rewrites, fallback providers, or unpublished tools. - This public skill is instruction-driven. Produce the artifact described by the workflow directly from the available evidence. - Do not call private provider/runtime paths or unpublished local tools. - If the CLI returns a quote-confirmation challenge, run `postplus quote confirm --json --challenge-file <challenge.json>` and retry with the returned token.