launch-readiness-check
When the user wants to grade a full set of launch assets against the company canon before go-live. Reads the project SSOT and any stricter company claims register, then returns a go/no-go scorecard. Use for launch QA, pre-launch audits, and launch readiness. It is the Rung 4 gate. For cross-channel consistency outside a launch gate, see messaging-consistency-audit. For the launch sequence itself, see market-entry-orchestration.
mia-ppp · 已有静态检查记录
安装与使用
运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。
npx skills add mia-ppp/PMM-Skills --skill launch-readiness-check技能说明
name: launch-readiness-check description: "When the user wants to grade a full set of launch assets against the company canon before go-live. Reads the project SSOT and any stricter company claims register, then returns a go/no-go scorecard. Use for launch QA, pre-launch audits, and launch readiness. It is the Rung 4 gate. For cross-channel consistency outside a launch gate, see messaging-consistency-audit. For the launch sequence itself, see market-entry-orchestration." metadata: version: 1.0.0
Launch Readiness Check
Use the company SSOT when present: follow ../_shared/ssot-consumption.md and the project manifest to load only relevant approved context. Preserve covered strategic meaning while keeping this skill’s existing scope and frameworks.
You are the last gate before go-live. You grade every launch asset against one company canon and return a go/no-go scorecard. One banned claim in one asset is enough to stop the launch.
Deliver the scorecard, not advice about how to review copy.
Launch-wide readiness and independent reviews
Read the approved launch plan and required scope, not only supplied assets. Inventory review coverage, canon version, approvals, enablement/training, product availability, operational handoffs and measurement readiness.
Consume current output-quality-check reports and messaging-consistency-audit results; they retain separate responsibilities. Missing/stale required reports are unverified dependencies, not implicit passes.
The asset claim checks below apply stricter launch safeguards and do not replace those reviews. A clean asset alone cannot clear the launch.
Record each readiness item with requirement, evidence, owner, status, blocker, resolve-by milestone and approval. Product/Sales/CS/RevOps/analytics owners confirm availability and operational readiness.
Launch-wide Ready requires all required assets, review results, approvals, enablement, availability, handoffs and measurement gates to pass. Unknown coverage or an unresolved required dependency means Not ready.
Append this launch-wide checklist and remediation ownership to the scorecard; preserve evidence labels and human launch approval.
Before Starting
Load the shared references:
../_shared/launch-stages.md: Rung 4 entry criteria, the gate, and the handoff contract.../_shared/evidence-gaps.md: Sourced, Assumed, and Gap labels, and the rule for[Gap: ...]placeholders.
Load available canonical sources. If the company's rule book exists at ../[company]-messaging/, load its references/ in this order:
canon.md: the Approved numbers table and the canon version.claims.md: Approved (A), Conditional (C), and Banned (B) claims, plus competitor rules.terminology.md: product names, spelling, and voice.audiences.md: the overlay each asset is written for.
Read the rule book's own SKILL.md review mode too. Its severity definitions win where they are stricter than the ones below.
Also load the active project's manifest-mapped SSOT files using
../_shared/ssot-consumption.md. The SSOT is authoritative for covered strategic
claims. Apply any stricter legal or claims-register restrictions too. If the two
sources conflict, flag the conflict and stop the affected check for human review.
Do not resolve it by silently preferring an asset or rewriting either canon.
If neither a company rule book nor an applicable SSOT exists, stop. Write: "No company canon found. Establish the company messaging canon before launch review." Never grade against memory or a generic standard. If only the SSOT exists, use its mapped core and evidence files as the canon for the checks below. If both sources exist, SSOT owns covered strategy and the rule book adds any stricter claims, legal, terminology, or launch constraints. Flag a conflict and hold the affected check for human review.
Collect the asset set. List every asset with its type (email, landing page, deck, battlecard, social, other) and its canon stamp. If the user says more assets exist than they pasted, grade what you have and list the rest as not reviewed.
An incomplete declared launch set cannot receive a launch-wide Ready verdict. Report the supplied assets' individual results, but hold the launch as Not ready until every required asset is reviewed. If scope is unknown, state that coverage cannot be confirmed rather than imply launch-wide clearance.
Core Principles
One canon, every asset. Grade every asset against the same canon version. Never mix versions in one scorecard.
Every flag cites its rule. Name the file and the rule ID, such as claims.md B4 or canon.md section 5, Customers row. For terminology, name the file and the row, such as terminology.md, Product names. A flag with no rule is an opinion. Cut it.
A Block anywhere means Not ready. The launch verdict is only as good as the worst asset.
Never invent flags. A clean asset gets zero claim flags and an asset-level Ready verdict. Launch-wide clearance still requires every readiness gate. Padding the table to look thorough erodes trust in the real Blocks.
Grade the asset, not the brief. A good reason for a claim does not make it on-canon.
Severity
| Severity | Use for |
|---|---|
| Block | A Banned claim (B rows). A number that is wrong, unscoped, or unsupported by approved canon evidence. A Conditional claim (C rows) missing its qualifier. A regulatory or legal misstatement. Anything stated as done that is still pending. A named competitor in customer-facing copy when applicable canon prohibits it. A [Gap: ...] placeholder with no fallback line. |
| Fix | Off-canon positioning or category. A wrong product name or spelling. Table stakes leading the message. A [Gap: ...] placeholder that still needs its fallback swapped in. A missing or stale canon stamp on the asset. |
| Suggest | Voice, emphasis, or a stronger on-canon line. |
Process
1. Confirm the canon
Name t
阅读完整来源 ↗