The market behind the project.
Five files — competitor landscape, market intelligence, persona pricing, whitespace atlas, website-design audit — plus one compiled brief. Produced by the competitor-intel orchestra (Agents 1–6). Each file is the rendered view of one JSON in /data/; the brief is what they compile to.
Internal desk — refreshed research date pending.
Five analyses + one compiled report.
Each file is produced by a single agent in the orchestra and writes to a single JSON in /data/. Click through to the rendered view; the underlying data is what the brief compiles.
Top-5 with rationale; full database searchable + filterable on category, region, threat. Agent 1's output.
Bottom-up market sizing with stacks and filters. Five personas with NBA arithmetic and elasticity bands. Tier ladder anchored to the buyer's NBA.
Where the product breaks the cluster, where the heatmap is empty, and named niches with TAM and beatability scores.
Five-dimension design rubric. Mobile pass. Three-intent findability test. Issues scored and prioritised. Agent 5's output.
Executive summary, market, competition, pricing, whitespace, design audit, recommendations, methodology & sources. Pre-flight gated against the inter-agent quality contract.
What the intel says — in four findings.
Top four implications from market-intelligence.json → market_size.implications[]. Each carries an agent_targets[] field that says which downstream agents the implication shapes.
Findings will appear once Agent 2 (market-intelligence-analyst) writes implications.
This dashboard reads the top four entries from market_size.implications[] and renders them as headline + body. Make sure Agent 2 produces structured implications, not bullet text.
Who to watch most closely.
Pulled from the top_five rationale field of the competitor file. The first two entries often carry the silent NBAs (DIY alternatives, in-house build) — note them as the unmarked baseline the named competitors fight against.
| Rank | Competitor | Why they matter |
|---|---|---|
| — | Top-5 will appear once Agent 1 writes top_five[]. | The rationale field is rendered here verbatim. Keep it ≤ 200 chars; explain WHY the competitor matters, not what they sell. |
Intelligence decays — keep it dated.
This desk is a snapshot, not a feed. Every figure is tied to a research date and a named source so a reader always knows how stale it is. See methodology/07-methodology-curator.md.
Competitor & design rescan
Re-run Agents 1 and 5 every quarter — competitor positioning and websites change fastest.
Market & pricing refresh
Re-run Agents 2 and 3 whenever a regulatory dependency in market.policies[] updates. The implications cascade into pricing tiers.
Full orchestra rerun
Rebuild TAM/SAM/SOM and the strategy canvas from scratch every May. Set meta.research_date on every JSON when it lands.