---
name: aeoforge-content-creation
description: Use when creating, auditing, comparing, refreshing, or clustering content optimized for AI answer engines like Google SGE, Perplexity, or ChatGPT. Triggers on AEO, answer engine optimization, content scoring, featured snippet extraction, JSON-LD generation, or keyword clustering tasks.
---

## Overview

AEOForge is a 25-tool MCP/API platform for answer-engine-optimized content.

## Mandatory First Steps (do these BEFORE any tool call)

1. Call `aeo_account` (or `GET /api/v1/me`) — verify auth, check credits and tier
2. Call `GET /api/v1/onboard` — receive full product context including:
   - Complete tool catalog with costs and descriptions
   - 8-dimension scoring system with weights, grade scale, and tips
   - Recommended workflow (research → write → score → iterate → schema → save)
   - Response header reference (X-AEO-Score, X-AEO-Grade, X-AEO-Weakest, etc.)
3. **Read the onboard response fully** — it contains scoring rules that directly affect article quality (e.g. Entity dimension requires passing research entities to /score, Schema dimension requires calling /schema then re-scoring)

**For API key callers (agents):** The `/api/v1/write`, `/api/v1/outline`, `/api/v1/polish`, and `/api/v1/generate` endpoints return 403. Your model writes better content than our GPT pipeline. Use `/api/v1/research` for intelligence, write the article yourself, then `/api/v1/score` for feedback.

## When to Use

- Creating new AEO-optimized content (research → outline → write → verify)
- Auditing existing pages for AEO readiness
- Comparing two articles for competitive analysis
- Refreshing stale or outdated content
- Building keyword strategy via clustering
- Formatting content for featured snippets, knowledge panels, PAA
- Generating JSON-LD schema markup (Article, FAQPage, HowTo)

## Quick Reference

### Create Phase

| Tool | Credits | Purpose |
|------|---------|---------|
| `aeo_research` | 10/15 | Live web research — standard (10cr) or deep mode (15cr, gpt-4o) |
| `aeo_outline` | 3 | Structured outline with snippet targets from research |
| `aeo_write` | 8 | Section-by-section draft with verified citations |
| `aeo_polish` | 2 | Style pass with brand voice, changelog, idempotent |
| `aeo_verify` | 2 | Fact verification with confidence tiers |

### Analyze Phase

| Tool | Credits | Purpose |
|------|---------|---------|
| `aeo_score` | Free | 8-dimension AEO score with grade (A+/A/B/C/D/F) and actionable suggestions |
| `aeo_schema` | Free | Article + FAQPage + HowTo JSON-LD generation |
| `aeo_extract` | Free | Featured snippet, knowledge panel, PAA formatting |
| `aeo_audit` | 5 | Fetch URL, score it, get prioritized improvement plan |
| `aeo_compare` | 3 | Head-to-head comparison across 7 AEO dimensions |
| `aeo_client_profile` | 5 | Build client intelligence profile (brand voice + company knowledge) |
| `aeo_optimise` | 8 | Page-level section analysis, intent classification, and AEO-optimised rewrites |
| `aeo_complete_audit` | Free | Comprehensive site-wide audit: discover pages, score, benchmark vs competitors |

### Maintain Phase

| Tool | Credits | Purpose |
|------|---------|---------|
| `aeo_refresh` | 8 | Fresh research + targeted update plan for stale content |
| `aeo_cluster` | 5 | Keyword clustering with pillar hierarchy + content calendar |

### Pipeline

| Tool | Credits | Purpose |
|------|---------|---------|
| `aeo_account` | Free | Check credits, tier, and per-tool cost table |
| `aeo_full_pipeline` | 12/20 | Draft (12cr) or publish (20cr) end-to-end pipeline |

### Workflows (Composite)
| Tool | Credits | Purpose |
|------|---------|---------|
| `aeo_workflow_audit_improve` | 5-13 | Audit URL + conditional refresh in one call |
| `aeo_workflow_strategy` | ~44 | Cluster → research → outline top N pillars |
| `aeo_workflow_health` | 5/URL | Batch-audit up to 10 URLs, priority dashboard |
| `aeo_workflow_quality_gate` | Free | Score pass/fail against target threshold |

### Competitive Intelligence

| Tool | Credits | Purpose |
|------|---------|---------|
| `aeo_crawlability` | Free | Check AI bot access: robots.txt for 11 AI bots, meta noai, X-Robots-Tag, llms.txt, sitemap, JS rendering |
| `aeo_llms_txt` | 3 | Generate spec-compliant llms.txt + llms-full.txt with semantic page grouping |
| `aeo_share_of_voice` | 10 | Measure AI answer visibility: SOV %, coverage, avg position, gap keywords across 3 engines |

## Core Patterns

- Always run `aeo_research` before `aeo_outline` — outlines without research lack competitive context
- Always run `aeo_verify` after `aeo_write` — catches ungrounded claims before publishing
- Use `aeo_score` (free) after every edit to track improvement
- Check `content_is_fresh` from `aeo_refresh` before doing unnecessary work
- Read `hallucinatedKeywords` from `aeo_cluster` to flag LLM-invented keywords
- Read `scoreBreakdown` from `aeo_extract` to identify weak extraction formats
- Pass research data between tools to avoid re-fetching
- Prefer composite workflows (`aeo_workflow_*`) over manually chaining individual tools when the workflow matches

## Error Handling

| Code | Meaning | Action |
|------|---------|--------|
| 401 | Invalid or missing API key | Check key starts with `aeo_sk_` |
| 402 | Insufficient credits | Call `aeo_account`, upgrade at /pricing |
| 429 | Rate limited | Wait `retry_after_seconds`, backoff |
| timeout | Tool exceeded limit | Retry once; reduce input size |
| fetch_failure | URL unreachable | Verify URL, try markdown input instead |

## Common Mistakes

- Skipping research before outline — produces generic, non-competitive content
- Sending raw HTML instead of markdown — tools expect clean markdown
- Ignoring `researchAvailable: false` on audit — competitive data unavailable
- Not verifying after writing — risks publishing hallucinated claims
- Using deep research (15cr) when standard (10cr) suffices
- Sending >200 keywords to cluster — capped at 200, sweet spot is 20-100