# Autonomous Content Pipeline

> Research, outline, draft, verify, edit, and preflight a batch of sourced content pieces by composing the catalog's atomic research and content skills, unattended. Use when a user has supplied a topic list, audience, objective, and publication requirements and wants review-ready drafts produced without manual handoffs between steps; this workflow never publishes.

- Skill: `knuckles-team/autonomous-content-pipeline` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add knuckles-team/autonomous-content-pipeline`
- Raw SKILL.md: https://api.skillmd.com/api/skills/knuckles-team/autonomous-content-pipeline/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- License: MIT
- Author: Knuckles-Team (https://skillmd.com/u/knuckles-team)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/knuckles-team/autonomous-content-pipeline

---


# Autonomous Content Pipeline Workflow

Compose the named atomic skills without adding editorial or publication logic
here.

## Inputs

Provide the topic list, audience, objective, channel and length constraints,
voice guide, citation style, and publication requirements.

## Steps

### Step 0: web-search [skill: web-search]

Invoke `$web-search` with the workflow inputs for each topic.

Expected: `source_packet`

### Step 1: content-outline-builder [skill: content-outline-builder] [depends_on: Step 0]

Invoke `$content-outline-builder` with the workflow inputs and
`source_packet`.

Expected: `approved_outline`

### Step 2: content-draft-writer [skill: content-draft-writer] [depends_on: Step 1]

Invoke `$content-draft-writer` with the workflow inputs,
`source_packet`, and `approved_outline`.

Expected: `sourced_draft`

### Step 3: citation-auditor [skill: citation-auditor] [depends_on: Step 2]

Invoke `$citation-auditor` with `sourced_draft` and `source_packet`.

Expected: `citation_audit`

### Step 4: copy-editor [skill: copy-editor] [depends_on: Step 3]

Invoke `$copy-editor` with `sourced_draft`, `citation_audit`, and the
workflow inputs.

Expected: `edited_draft`

### Step 5: publication-preflight [skill: publication-preflight] [depends_on: Step 4]

Invoke `$publication-preflight` with `edited_draft`,
`citation_audit`, and the publication requirements.

Expected: `readiness_decision`

## Output

Return `edited_draft`, `citation_audit`, and `readiness_decision` for each
topic. Do not publish, schedule, send, or upload any piece.

## Execution

- **Run first:** Step 0 — `$web-search`.
- **After level 0:** Step 1 — `$content-outline-builder`.
- **After level 1:** Step 2 — `$content-draft-writer`.
- **After level 2:** Step 3 — `$citation-auditor`.
- **After level 3:** Step 4 — `$copy-editor`.
- **After level 4:** Step 5 — `$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.