Roadmap

The original 9-tool build plan is complete — see the archived build plan for what shipped and ADR-0003 for how it was consolidated into one deployment. This page is the forward-looking roadmap.

GitHub issues are the source of truth. An item here is a theme; the issue is where the design gets discussed and where you claim the work. If an item has no issue yet, open one before starting. Ordering within a section is rough priority, not commitment.

Dogfooding the product on our own site

From the SEO + AEO plan — the toolkit teaches these patterns, so its own site has to ship them:

• Glossary pages (/glossary/*) — definition pages for AEO/GEO terms (AEO, GEO, llms.txt, E-E-A-T, answer engine, citation), interlinked into a topic cluster. Definitional queries are where AI assistants cite most readily.
• Comparison pages (/compare/*) — starting with /compare/aeo-vs-seo; the comparison/listicle queries answer engines synthesize from.
• Cornerstone guide (/guide/answer-engine-optimization) — the what/why/how pillar page, linking to every tool.
• Self-audit in CI — run our own auditor against advancelabs.dev in a scheduled workflow and fail below a threshold, so the site can never quietly regress on the rules it sells.
• Vercel Speed Insights — close the Core Web Vitals measurement gap in the plan’s §7.

Scoring engine

• Starter rules — #10 tech.charset-declared, #11 aeo.content-freshness, #12 tech.hreflang-valid.
• Opportunity scoring — rank findings by expected impact rather than severity (#30).
• The rule set is never finished — see Adding a scoring rule in CONTRIBUTING for the recipe; a new rule moves the pinned counts in rules.test.ts and the published copy that quotes them.

Search-data tools (GA4 + GSC)

• Topical hub mapping — cluster GSC queries into topic hubs (#31).
• GSC history retention — persist snapshots to break the 16-month API ceiling (#29).

Backlinks

• Optional paid data provider — DataForSEO behind the existing backlinks seam, BYOK as always (#20).

Infrastructure & code health

• Postgres token store — durable tokens without requiring Supabase (#23).
• Split AuditExperience.tsx — 1061 lines into focused components (#14).
• Test coverage for @advance-labs/llm json helpers (#13).

Managed tier

Tracked as TODO(lead) markers in the code rather than issues, since they depend on production configuration: live delivery counts and a citation snapshot in the account panel, a public route for the guarantee terms, and a citation-coverage snapshot at onboarding. The activation runbook covers turning the tier on.

Out of scope here

AI-visibility tracking over time (are the engines actually citing you, week over week) deliberately lives outside this repo — see visibility tracking for why and where.