Scaffold Kit: A Clear Plan Before Your AI Builds

An AI can start writing code before you have agreed on what it should build. Scaffold Kit helps you make those decisions first.

What it does

It guides your assistant through the questions that matter, explains meaningful tradeoffs and produces four linked project documents:

• Architecture Document: what you are making and how its parts fit together.
• Engineering Document: requirements, data, risks and checks for acceptable work.
• Decision Log: choices, alternatives, unanswered questions and reasons to reconsider.
• Slice Brief: the next bounded piece of work, its exclusions and how success will be checked.

Choose plain language, balanced explanations or developer shorthand. The explanation changes; evidence and approval requirements stay in place. Unanswered choices remain visible rather than becoming invented answers.

Use it to plan a new project, organize an existing one, or define the next feature. Small changes use a compact brief instead of requiring another full planning interview.

Works independently

Anti-dark-code is optional. Scaffold Kit can inspect the relevant existing code directly and use your project's tests or recorded manual checks. If you use both skills, it can link to Anti-dark-code's findings and verification evidence without duplicating them.

Optional bundle: Scaffold Kit + Anti-Dark-Code: Plan and Verify is $39 one-time, saving $9 against $48 separately. Both skills remain available individually.

Requirements

• Your own compatible AI agent, with access to the skill's supporting references/templates and permission to read relevant project files and write planning documents.
• Your input on goals, consequential choices and approval of the specific build brief.
• Python 3.9+ only for the optional mechanical document auditor; it needs no additional Python packages.
• An AI account and any model/API usage are separate from this purchase.

Limitations

• This skill plans work; implementation requires your separate build approval.
• A passing document audit checks structure and consistency. It does not establish approval, passing software tests or human comprehension.
• Results depend on the information and project evidence available. Missing facts and contradictions should be reported rather than filled with plausible explanations.
• Supporting-file and script features depend on host capabilities. In Claude, creator-free Agensi access retrieved the main instructions and six supporting text files: all four planning templates, the planning conductor and the auditor source.
• These checks do not cover buyer checkout/unlock or delivery, running Agensi-delivered helpers, every remaining file or conditional workflow, or complete byte-for-byte matching to the submitted ZIP.
• The example screenshots came from the exact creator-supplied ZIP in separate Claude conversations; they do not demonstrate buyer delivery.

Source and version

Based on the 0.4.4 source with the interoperability fixes merged in PR #11. This submission is not a newly published GitHub release.

The source is publicly available on GitHub and includes FSL-1.1-MIT license terms. The listing offers the packaged workflow through Agensi. https://github.com/LynxTWO/scaffold-kit