Everything IdeaNavigator AI does, explained.

Verdict (Build / Validate / Research / Rethink), 0–100 score, confidence, and this-week's test up top.

A 5-criterion weighted rubric scores every report; see the Validation section below.

A dated day-1→day-7 plan to test the riskiest assumption cheaply.

Derived deterministically from the report's buyer, evidence, and validation test.

Price-anchored revenue scenarios (proof / wedge / vertical-leader) and a tiered offer.

Computed from the report's monetization model and buyer.

Segments, channels, intent keywords, messaging angles, and objections per report (HTML + .md, per idea at /audience/<slug>/).

Derived from buyer, problem, and community signals.

Time-to-execute score, bottlenecks, accelerators, and a dated launch plan (HTML + .md + .ics).

A readiness model over the execution plan; the launch plan exports to your calendar.

The report's links in the intelligence graph: complaint evidence, trend seed, adjacent verticals, sibling reports — each with provenance.

Precomputed from the ontology graph; declared evidence is shown before inferred.

Incumbents, substitutes, whitespace, plus public companies/funding the graph links to the vertical.

Prose analysis joined with the company/funding signal lane.

A stage (Emerging→Validating→Heating→Crowding→Cooling) and a 0–100 market-timing score.

Deterministic rules over re-check trajectory, demand, saturation, and complaint decay.

Adjacent wedges, founder-fit scoring, and a hard-questions 'roast' of blind spots.

All derived from the report's structured frontmatter.

Each report as .json, .md, .ics (calendar), backlog.json (GitHub/Linear issues), and a print/PDF dossier.

Static endpoints + a chromium print-to-PDF step in the pipeline.

An interactive force-directed map of the whole ontology; click a node for its neighborhood and edge provenance.

Client-side render over /graph.json with a deterministic seed layout.

A browsable page per non-report node — clusters, trends, tools, companies, funding, adoption signals, regulations, job roles, personas, keywords, communities.

Generated for every entity with enough links to be worth a page.

The mechanical, entity-level diff ledger: what was added, updated, or removed, filterable by vertical and entity type.

A daily fingerprint of the graph; corruption-guarded and idempotent.

An independence score per report: how many distinct source domains back it, the declared/classified/inferred mix, and a single-source-family flag.

Aggregates per-edge provenance across all evidence lanes.

Weekly time-series per vertical from the change ledger — which industries are gaining over time.

ISO-week aggregation with sparklines; richer every day.

How each publish-week's ideas age as a group (30/60/90-day survival).

Replays the ledger + re-check signals at each age band.

The full graph: typed entities + provenance-stamped edges (INAV-GRAPH-1).

Static endpoint; the same data behind every graph surface and the MCP graph tools.

A short quiz that sets your founder type and feeds recommendations.

Stored in localStorage; merged into the For-You score.

Scores a specific report against your skills, access, time, capital, and sales comfort.

Client-side scoring over your profile.

Match markets/skills/budget/style to a research handoff.

Local matcher over the archive.

Your saved-idea profile and a side-by-side comparison table.

localStorage-backed; exportable.

Send a saved shortlist to a co-founder — the whole list is encoded in the URL.

A base64url codec; the recipient renders cards from /ideas.json, no backend.

Watch verticals, entity types, or saved searches; see what moved since you last looked.

Diffs /changes.json against your last-seen, locally.

A personalized 'what changed for your thesis' combining monitors, momentum, and fresh reports.

Assembled client-side from your watchlist.

Clustered user complaints from public developer sources.

Mined daily, clustered, and linked to reports as declared evidence.

Live trending terms that seed slot-2 daily ideas.

Fetched from the Stenvrik bridge; topic-gated.

A weekly snapshot of the AI assistant/tooling market.

Generated weekly; tools enter the graph.

Public competitors and funding rounds, linked to verticals and reports.

Curated seed + fail-soft fetcher; powers the 'who's moving here' panel.

Real npm/PyPI/GitHub download & star slopes — a quantitative demand signal.

Dated time-series appended daily; feeds momentum and thesis confidence.

Public job-posting volume as a budget-confirmed demand proxy per vertical.

Fail-soft fetch; corroborates persona entities.

Future compliance deadlines whose urgency grows as the date nears.

Federal-register-style records; drives 'why now' and monitor alerts.

Per-vertical public sources feeding under-served industries (healthcare, education, government, nonprofits).

A source-pack framework merged into the complaint flow.

Read-only tools for ideas, verticals, the graph, changes, provenance, thesis scoring, brief drafting, backlog, distribution, snapshot verification, and more.

Ships in the repo; consumes the public JSON, no key.

ideas, verticals, graph, changes, themes, collections, weekly, distribution, search-index, score-rubric, metrics, per-idea .json/.md/.ics/backlog.

All statically generated at publish time.

Cryptographic SHA-256 hashes of every canonical endpoint + a top digest (INAV-MANIFEST-1).

Lets a consumer pin a snapshot and detect drift; ideanavigator_verify_snapshot.

Register a target + monitor selection and get signed POSTs when matching changes land.

PHP intake + a Mac-mini HMAC-signed fan-out, with SSRF protection on targets.

An agent-orientation file and site + per-vertical RSS feeds.

Static; llms.txt indexes capabilities, URLs, and conventions.