geoptimize 0.9.0

What's new

• Offline visual reports keep the original five readiness dimensions alongside site health charts, page search and expandable evidence.
• Site metrics expose 19 observed counts with bounded before/after comparisons.
• Optional scan --details records all 17 rule results and source excerpts with capture provenance. Reports can compare compatible readiness scans and show matched source excerpts.

The scoring algorithm and default scan JSON remain unchanged. Detailed source capture is opt-in.

Install

npm install -g geoptimize@0.9.0

geo scan ./dist --dir --details --json > readiness.json
geo report readiness.json --output report.html

Add --baseline previous.json for readiness comparisons and --site site-audit.json for website health charts. Scans must have matching versions and page sets for readiness deltas.

geoptimize v0.8.1: aeoptimize migration

geoptimize 0.8.1 publishes the aeoptimize project lineage and migration guide on npm.

The package continues the same project formerly published as aeoptimize. Existing geoptimize API and CLI names and runtime behavior are preserved.

geoptimize on npm · Legacy aeoptimize package · aeoptimize 0.7.1

geoptimize v0.8.0

Formerly aeoptimize. geoptimize continues the same project under a new npm name. The legacy aeoptimize package and v0.7.0 release retain the earlier package history. Follow the migration guide to update existing integrations.

geoptimize 0.8.0 moves the package, CLI, framework integrations, and GitHub Action to the GEO naming. Deterministic scoring and audit evidence contracts are preserved.

Migration

• Install geoptimize@0.8.0 and use geoptimize, geo, or geo-cli.
• Update imports to geoptimize, geoptimize/vite, or geoptimize/next; use geoPlugin and withGeo.
• Update bundled skill references to geo-* and generated schema paths to _geo.
• Pin the GitHub Action to cucuwang/geoptimize@v0.8.0.

- uses: cucuwang/geoptimize@v0.8.0
  with:
    path: dist

The Action is advisory by default. Set fail-on-low-score: 'true' and a project-specific min-score for a blocking check.

Updated js-yaml and Vitest dependencies address published security advisories.

aeoptimize v0.7.1: geoptimize migration

aeoptimize 0.7.1 links the legacy npm package to the continuing geoptimize project and migration guide.

The aeoptimize, aeo, and aeo-cli commands, aeoPlugin, and withAeo remain available. This patch adds the migration notice and refreshes js-yaml and Vitest lockfile entries for the current security audit. The legacy package remains installable and is not deprecated.

aeoptimize on npm · geoptimize on npm

v0.7.0

This project continues as geoptimize from v0.8.0. This release belongs to the original aeoptimize npm package. See the migration guide and geoptimize v0.8.0 release.

aeoptimize 0.7.0 adds evidence-backed audits for individual pages, built HTML, and bounded same-origin site crawls.

• audit reports page-level observations with remediation and validation guidance.
• audit-build finds malformed JSON-LD, invalid canonical declarations and repeated page metadata, with optional --expect-indexable --fail-on-error CI gates.
• audit-site checks discovered pages, internal-link targets, canonical conflicts and sitemap discrepancies within an explicit crawl limit.

The v0.6 scoring methodology and existing scan --json contract remain unchanged. Unavailable evidence is reported separately from confirmed failures.

npm install --save-dev aeoptimize@0.7.0
npx aeoptimize audit-build ./dist --json

Release validation passed with 208 tests, Node.js 22 and 24 CI, Action contract checks, zero reported vulnerabilities, and byte-identical package manifests across local and CI builds. The source commit is cc7345c8782ce7e691df328c9cdc5e9e686adfb8.

aeoptimize v0.6.2

aeoptimize v0.6.2

Provenance-corrective release

• Publish the verified v0.6 candidate from the exact merged main commit.
• Add a fail-closed prepublish source gate requiring HEAD to match the freshly fetched origin/main.
• Preserve the artifact-bound public verification of npm metadata, tarball SHA-256, CLI aliases, Git tag, and GitHub Release state.

The npm 0.6.1 package contained the verified candidate contents, but its gitHead recorded the pre-squash PR commit rather than the merged commit. No v0.6.1 Git tag or GitHub Release was created. Version 0.6.2 fixes forward with matching npm gitHead, Git tag, and merged commit provenance.

Verified artifact

• npm package: aeoptimize@0.6.2
• release commit: 3b8593802595771ac40d1993a44e682ae5a1bdfc
• tarball SHA-256: c25a5c9daaee9f3c8c0fc412a794c8353dc14a38b70c8cd5a2da53c1ccbf28ed
• CLI aliases: aeoptimize, aeo, and aeo-cli
• CI: Node.js 22 and 24 candidates were byte-identical

No scoring rule, rule weight, JSON field, Action input, Action output, or runtime behavior changes in this patch. This release makes no ranking, traffic, indexing, rich-result, AI Overview, or citation outcome claims.

aeoptimize v0.6.0

aeoptimize v0.6.0

Evidence-bounded readiness scoring

• Reframe the score as deterministic content readiness rather than a prediction of ranking or AI citation.
• Treat FAQ structure and llms.txt as optional, zero-point signals.
• Remove exact-one-H1 and fixed meta-description-length assumptions.
• Flag unsourced quantitative claims instead of rewarding more numbers.
• Publish a versioned fixture corpus with positive, negative, and false-positive boundaries for every scored rule.

Reproducible automation

• Support maintained Node.js 22 and 24 lines.
• Preserve aeoptimize, aeo, and aeo-cli in the packed npm manifest.
• Add an advisory-by-default GitHub Action with explicit blocking mode and stable JSON outputs.
• Add a copyable end-to-end Action sample and release-time contract checks.
• Refresh dependencies and require zero high or critical audit findings at release time.

This release does not claim ranking, traffic, indexing, rich-result, AI Overview, or citation outcomes. The score is intended for reproducible regression checks within the same project, aeoptimize version, configuration, and fixture set.

v0.2.0 — Multi-AI Scoring

What's New

Multi-AI Scoring Engine

Score your website's AI readiness with multiple AI engines simultaneously.

npx aeoptimize scan https://your-site.com --multi-ai

How it works:

• Detects available AI CLIs (gemini, copilot) on your system
• Dispatches parallel scoring requests to each
• Merges with rule engine using weighted average
• Falls back gracefully to rule engine only if no CLIs found

Scoring methodology:

Scenario	Weighting
Rule engine + 2+ AIs	50% rules + 50% AI average
Rule engine + 1 AI	60% rules + 40% AI
Rule engine only	100% rules

Example output:

Score: 72/100 (Rule Engine: 61 | AI Consensus: 83)

  Rule Engine      ████████████░░░░░░░░ 61/100
  Claude           ████████████████░░░░ 85/100
  Gemini           ████████████████░░░░ 81/100

AI Insights:
  Claude:  "Content is well-structured but FAQ section lacks schema markup"
  Gemini:  "High citability, missing llms.txt reduces discoverability"

New Files

• ai-prompt.ts — Standardized scoring prompt for all AI engines
• external-scorers.ts — CLI detection and invocation
• merger.ts — Score aggregation logic
• aeo-ai-scorer agent — Claude Code subagent for semantic scoring

Claude Code Skill Updates

• /aeo-scan now detects available AI CLIs and offers multi-AI scoring
• New aeo-ai-scorer agent for Claude-based semantic analysis

Stats

• 50 tests passing
• 17 rule engine rules + multi-AI scoring
• Zero external dependencies for AI scoring (uses existing CLI subscriptions)

v0.1.0 — Initial Release

aeo-cli v0.1.0

First release of aeo-cli — AI search optimization toolkit.

Features

CLI (npx aeo-cli)

• aeo scan <url|dir> — AI readability audit (0-100 score across 5 dimensions)
• aeo generate <dir> — Generate llms.txt, JSON-LD schemas, robots.txt suggestions
• --json flag for machine-readable output

Claude Code Skills

• /aeo-scan — Interactive audit with discussion and fix suggestions
• /aeo-generate — Guided AI infrastructure file generation with preview
• /aeo-transform — LLM-driven content restructuring (zero extra cost)

Scoring Engine

• 17 rules across 5 dimensions: Structure, Citability, Schema, AI Metadata, Content Density
• Pure rule-based (offline, zero cost, deterministic)

Install

# CLI
npx aeo-cli scan https://your-site.com

# Claude Code skill
claude plugin marketplace add dexuwang627-cloud/aeo-cli