Skip to content

CLAUDE.md

Latest commit

 

History

History
110 lines (82 loc) · 3.74 KB

CLAUDE.md

File metadata and controls

110 lines (82 loc) · 3.74 KB
Raw

FLOW — Claude Code Integration Guide

FLOW is an evidence-led SEO knowledge base designed for direct consumption by AI agents. This file configures how Claude Code and SEO-Agent interact with this repo.

What This Repo Is

FLOW (Find · Leverage · Optimize · Win) is a four-stage SEO playbook with:

  • 72 docs covering the full SEO lifecycle
  • 42 AI-ready prompts in docs/09-prompts/ organized by stage
  • 27 diagrams as assets
  • llms.txt + llms-full.txt per the llmstxt.org spec for agent ingestion
  • Every claim anchored to a 2026 source with retrieval date

Primary AI Consumer: SEO-Agent

SEO-Agent ingests this repo into its brain at session start:

# Mirror docs/ into wiki/flow/ and ingest into brain
python claude-seo/scripts/sync_flow_content.py --source /path/to/flow

# Or ingest directly if already mirrored
python -c "
import sys; sys.path.insert(0, 'claude-seo')
from pathlib import Path
from seo_agent.brain import AgentBrain, ingest_markdown_directory
from seo_agent.core.paths import resolve_paths
paths = resolve_paths()
brain = AgentBrain(paths.brain_dir)
brain.initialize()
count = ingest_markdown_directory(brain, Path('wiki/flow'), source_prefix='flow')
print(f'Ingested {count} files')
"

Once ingested, every SEO-Agent REPL session injects relevant FLOW doctrine into the system prompt via brain.context_block(query). You can also query directly:

seo-agent> /recall keyword research

Quick Start for Other AI Agents

Point any agent at llms-full.txt for one-shot ingestion of the full corpus:

# Claude Code / Cursor / Codex / Gemini
Read the file llms-full.txt for complete FLOW doctrine context.

Or use llms.txt for the short index and fetch individual docs on demand.

Directory Map

Path Contents
docs/00-START-HERE.md Entry point — what FLOW is and how to use it
docs/01-framework/ FLOW loop, AI search surface map, independence rules
docs/02-foundations/ Search intent, E-E-A-T, technical + schema foundations
docs/03-find/ Keyword research, audience avatar, query fan-out
docs/04-leverage/ Distributed presence, local citations, Reddit/LinkedIn
docs/05-optimize/ GEO formatting, entity consistency, content CTR
docs/06-win/ BOFU, dual-surface scorecard, PPC
docs/07-local-seo/ GBP, local citations, audits, local content
docs/08-playbooks/ SaaS, e-commerce, service business, B2B, affiliate
docs/09-prompts/ 42 stage-specific prompts ready for agent invocation
docs/10-references/ Bibliography, attribution, stats provenance

Prompt Library

Prompts in docs/09-prompts/ follow this pattern:

docs/09-prompts/
  find/        ← keyword research, query fan-out, audience avatar prompts
  leverage/    ← citation building, distributed presence prompts
  optimize/    ← entity, GEO, CTR optimization prompts
  win/         ← BOFU, scorecard, measurement prompts
  local/       ← GBP, local audit prompts

Each prompt is in a fenced text block — copy the content directly into your agent invocation.

Evidence Standard

All claims in this repo must cite a source with:

  • Publication year ≥ 2024
  • Retrieval date
  • Primary source (no AI-generated stats)

The CI script scripts/check_docs.py enforces this on every push. If you add docs, run it first:

python scripts/check_docs.py

SEO-Agent ↔ FLOW Sync

SEO-Agent mirrors docs/ into wiki/flow/ in its own repo. The sync is one-way:

FLOW repo docs/ → (sync_flow_content.py) → SEO-Agent wiki/flow/ → (ingest) → brain

To update after changes to this repo:

# From SEO-Agent repo root
python claude-seo/scripts/sync_flow_content.py