mcpbeat

Codex Slides Structured Intake

nexu-io/codex-slides-structured-intake

Collect and persist structured presentation requirements for Codex Slides, including scenario choice, audience, page count, language, format, research mode, visual direction, source-file roles, and clarification checkpoints. Use when a new deck request is ambiguous or scenario-led.

602 tokens
context cost
the whole folder, loaded on every use
2
files
instructions only
0
copies elsewhere
how many repositories repackaged it
748
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/nexu-io/codex-slides --skill codex-slides-structured-intake

The instruction itself

2 sections, as written by the author

Codex Slides Structured Intake

The Codex Slides Browser owns the canonical intake form. Structured intake must

produce one durable project and one shared clarification state, not a separate

chat-only questionnaire.

Flow

  • Call open_codex_slides without a project id and follow its

browserHandoff to the home/create workspace.

  • When a workflow preset fits, call list_scenarios, then open the selected

scenario deep link. Use product-owned defaults and source slots.

  • Submit the user's topic and available files in the Browser. Assign every

uploaded file its scenario role; do not flatten datasets, source decks,

brand assets, and visual references into one undifferentiated list.

  • Let the product create the durable project before clarification. Keep the

clarification checkpoint visible and persist answers in project workflow

state.

  • Ask only questions that materially change the result: audience, purpose,

page count, language, aspect, research depth, evidence constraints, and

visual direction. Use recommendations already present in the form.

  • Continue through outline and inspiration checkpoints in the same Browser

project. Do not call the unattended create_deck path unless the user asked

for a headless run.

If Browser submission is unavailable, use start_project as the deterministic

fallback, immediately follow its clarification handoff, and keep later choices

attached to that same project id. get_onboarding_questions may help formulate

the intake, but its output is not a second source of truth.

When a durable outline or render job is useful, call start_project_run, keep

its runId, and use wait_project_run or cancel_project_run without closing

the Browser workspace.

How to use it

Copy the folder

Take nexu-io/codex-slides-structured-intake from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.