Presentation Builder Workflow
Compose the named atomic skills without adding new slide-rendering logic here.
Inputs
Provide the topic, audience, objective, length constraints, and voice guide.
Steps
Step 0: content-outline-builder [skill: content-outline-builder]
Invoke $content-outline-builder with the workflow inputs.
Expected: approved_outline
Step 1: content-draft-writer [skill: content-draft-writer] [depends_on: Step 0]
Invoke $content-draft-writer with the workflow inputs and
approved_outline to draft the talk track.
Expected: talk_track_draft
Step 2: marp-presentations [skill: marp-presentations] [depends_on: Step 1]
Invoke $marp-presentations with talk_track_draft and
approved_outline to render the Marp slide deck.
Expected: slide_deck
Step 3: publication-preflight [skill: publication-preflight] [depends_on: Step 2]
Invoke $publication-preflight with slide_deck and the publication
requirements.
Expected: readiness_decision
Output
Return slide_deck and readiness_decision. Do not publish or present it.
Execution
- Run first: Step 0 —
$content-outline-builder. - After level 0: Step 1 —
$content-draft-writer. - After level 1: Step 2 —
$marp-presentations. - After level 2: Step 3 —
$publication-preflight.
Execution: If graph-os is reachable, offload the whole DAG via graph_orchestrate action=execute_workflow (or the kg-delegate skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet depends_on in parallel, then their dependents.