Pricing / decision notebookInformation

Package notes

Know what
you are using.

Pricing is an instruction-based SaaS strategy Skill for founders and product or growth teams reviewing a subscription offer. It helps an AI assistant draft and review the choices behind a price, then identify the evidence needed to validate those choices.

Where it fits

Use it to compare a charging metric, rethink subscription tiers, form an initial price hypothesis, plan a price change, or review the clarity of an actual pricing page. It is unsuitable for declaring an optimal price without evidence or automatically changing live billing.

Load the package in an assistant supporting Agent Skills, then paste or attach readable materials and state the decision you want reviewed. Your existing AI host supplies the model; its normal access and usage costs still apply even though supplied-material drafting needs no separate API key.

Version
Original Skill 2.1.2.
Input
User-supplied product, customer, feature, usage, pricing, and business information. Page text or HTML when auditing a pricing page.
Output
Metric comparisons, tier proposals, price hypotheses, assumption ledgers, research/rollout plans, or a two-axis pricing-page teardown.
Runtime
A host assistant that supports Agent Skills. Text drafting does not require a separate API key or a company-account connection.
Included
SKILL.md, four reference documents, eight eval specifications, README, provenance, NOTICE, and MIT License.
License
MIT. Copyright (c) 2025 Corey Haines. Original attribution is preserved.

Evidence requirements

Prices derived from incomplete data remain hypotheses. Actual price research, conversion, churn, margins, competitor information, and page evidence must come from supplied or verified sources.

Existing product-marketing context is read when present. The assistant should ask for missing business context rather than invent it.

Implementation boundary

The package includes instructions and references, not a billing integration, survey engine, model subscription, document reader, or export program. Host capabilities determine file reading and browsing.

Billing changes, page implementation, survey execution, and broader CRO need separate tools and authorization.

What has been demonstrated

One Codex instruction-guided task on October 11, 2026 turned a fictional StudioRoute briefing into a metric comparison, three-tier proposal, monthly/annual price hypotheses, an assumption ledger and a validation plan. The authoring assistant inspected the saved output and separately recomputed its annual-price arithmetic.

This shows a supplied-material drafting workflow. Installed-client activation, an independent evaluator, the eight-eval suite, a real pricing-page teardown, billing changes and commercial effectiveness were not tested. Benchmark and Showcase provide the detailed scope and artifact evidence.

Provenance

Organized from coreyhaines31/marketingskills, skills/pricing, fixed commit b9ba399dd88b082b926e261e8ccfb843d20aa066. Original Skill, references, eval definitions, and license were retained. Packaging documentation explains scope; it does not establish an upstream endorsement.