# Zoning Signal — Full Corpus > Last-Updated: 2026-04-29 A civic intelligence observatory for places in motion. ## What Zoning Signal Is Zoning Signal reads the planning records that shape a region's future and publishes them as a living atlas. Agendas, minutes, staff reports, comp-plan amendments, ordinance drafts, capital improvement plans — the documents that constitute the public regulatory field — are harvested, synthesized, and projected onto reading surfaces calibrated for the people who act on them: developers, investors, brokers, attorneys, business operators, residents, civic leaders. The civic record moves before the market reacts. A 0-5 denial of a gateway-parcel rezoning, a vote-flip on a comprehensive plan amendment, a quietly-funded road-widening that crosses three jurisdictions — these are the leading indicators of what a region is becoming. They are also the records that no commercial intelligence product currently reads end to end. Zoning Signal closes that gap. The observatory's first coverage area is south Lake County, Florida — the US-27 corridor through Clermont, Leesburg, Minneola, and Groveland, each with distinct planning instruments and converging on a shared infrastructure spine. The architecture is regional. The methodology generalizes to any planning jurisdiction with a public records pipeline. ## Built by MainThread Zoning Signal is one public specimen of MainThread's craft — the *Decoder* archetype, applied to the civic substrate. MainThread is a Possibility Space Engineering Studio building intelligence systems that read complex fields and surface what is becoming legible inside them. The studio's other surfaces apply the same craft to job markets, real-time games, and the studio's own design language. The intelligence pipeline (signal harvest, lens-projection synthesis, named-pattern recognition) lives in a sibling NLAA. The observatory is the publishing surface — the point at which the multi-projection intelligence becomes legible to the public web and to AI agents. ## Methodology ### Calibrated Attention Heads Every published artifact is the product of a multi-lens read. Seven analytical positions (developer, investor, broker, attorney, business operator, resident, civic leader) inhabit each meeting reading and each place dossier as distinct cognitive surfaces. The same evidence resolves differently to each position; the synthesis is the projection that holds all seven in superposition and names the dialectics. The lens model is enforced at authorship. A meeting reading is not a synthesis with audience filters applied — it is seven distinct projections plus a synthesis projection. Each projection is a deliberate, crafted artifact. The synthesis is written last, after the analyst has held every single-lens position. ### Multi-Lens Projection on the Web Surface The web surface ships every lens block in HTML so AI crawlers can ingest every cognitive position. The reading surface defaults to synthesis; readers narrow to a single lens through a switcher in the page chrome. Lens state persists across sessions. AI agents can also pre-route by calling the MCP tool surface with a `lens` parameter — see the MCP section below. ### Source Pipeline Three primary substrates feed the observatory: 1. **CivicClerk** — the agenda-management platform used by most south Lake municipalities. Provides agenda packets, minutes, and meeting recordings via predictable URL structures. 2. **Laserfiche** — the public-records archive used by Lake County and several constituent cities. Indexed differently per jurisdiction; harvest layer normalizes. 3. **Direct PDF** — the long-tail substrate. Smaller boards (community redevelopment agencies, technical review committees, advisory boards) often publish minutes as standalone PDFs without an indexing layer. Direct fetch + validation closes this gap. Every fetched document carries a SHA-256 fingerprint and an audit-log entry. Re-embedding the document corpus is checksum-keyed: only changed documents re-enter the embedding pipeline. The Voyage 4 1024-dimensional embedding space backs semantic search across the published corpus. ### Voice Discipline The project's voice is the MainThread voice constitution applied to civic intelligence. Every meaningful sentence runs the six-point diagnostic: operator-voice slip, coach-voice slip, cheerleader-voice slip, process-narration slip, superlative or weakener or generic verb, comparison or negation. The dominant anti-pattern in civic-tech and AI-generated copy is directorial narration — *it should feel like*, *the reader will experience*. The observatory never instructs the reader on how to feel; it produces the feeling through specificity. ## Page-Type Architecture Four canonical artifact types, each grounded in a specific intelligence object, each with a fixed slot in the URL grammar. ### Place Dossier — `/places/{city}-{state}` The city or jurisdiction view. Distilled from per-city synthesis in the NLAA. Sections: Plain-English Summary, Primary Forces, Recent Motions, Why It Matters (multi-lens), Watch Next (multi-lens), Source Trail. Schema: `Place` + `AdministrativeArea` + `GovernmentOrganization` + `NewsArticle` + `FAQPage`. ### Corridor — `/corridors/{name-slug}` The cross-municipal economic-topology view. Reads decisions across boundaries (Minneola + Clermont + Groveland + Leesburg share the US-27 spine, the same aquifer, overlapping developer networks). Sections: Corridor Thesis, Pressure Map, Municipal Crosscurrents, Infrastructure Unlocks, Commercial Gravity, Residential Expansion, Strategic Implications. Schema: `Place` + `GeoShape` + `NewsArticle`. ### Meeting Reading — `/meetings/{jurisdiction}-{type}-{yyyy}-{mm}` The temporal artifact. One per harvested meeting. Sections: Meeting Snapshot, Plain-English Summary, Signal Extraction, Items of Interest, Entity Map, What Changed (delta from prior), Why It Matters, Source Trail. Citation Scaffold structure (BLUF + statistic + attribution + structured list + FAQPage). Schema: `NewsArticle` + `Event` + `GovernmentOrganization` + `FAQPage`. ### Signal Brief — `/briefs/{title-slug}` The editorial layer. Named-pattern essays drawn from the regional synthesis. Field Notes cadence — one named pattern per essay, 400-1,200 words, weekly. Sections: The Signal, Why It Matters, The Evidence, The Pattern, Who Should Care, Watch Next, Source Trail. Schema: `NewsArticle` + `spatialCoverage` + `Speakable`. ### Navigational Surfaces Hubs: `/` (the field), `/field/{region}`, `/places`, `/corridors`, `/meetings`, `/briefs`, `/method`, `/sources`, `/built-by-mainthread`. Auth surfaces (gated, never gating content): `/login`, `/signup`, `/account/watchlists`, `/watchlists/{slug}`. ## Named Patterns Under Active Tracking - **The Self-Storage Canary** — anchored in [The Self-Storage Canary](https://zoningsignal.com/briefs/self-storage-canary) - **The Quiet Revolution** — anchored in [The Quiet Revolution on Highway 27](https://zoningsignal.com/briefs/quiet-revolution-highway-27) - **The Grandfather Window** — anchored in [The Grandfather Window](https://zoningsignal.com/briefs/grandfather-window) - **The Bellwether Gas Station** — anchored in [The Bellwether Gas Station](https://zoningsignal.com/briefs/bellwether-gas-station) - **The Six-Month Board Flip** — anchored in [The Six-Month Board Flip](https://zoningsignal.com/briefs/six-month-board-flip) ## Published Place Dossiers - **[Groveland, Florida](https://zoningsignal.com/places/groveland-florida)** — Florida — Signal: Elevated · Rising - **[Clermont, Florida](https://zoningsignal.com/places/clermont-florida)** — Florida — Signal: Elevated · Accelerating - **[Leesburg, Florida](https://zoningsignal.com/places/leesburg-florida)** — Florida — Signal: Elevated · Stable - **[Minneola, Florida](https://zoningsignal.com/places/minneola-florida)** — Florida — Signal: Elevated · Accelerating ## Published Corridors - **[US-27 South Lake Corridor](https://zoningsignal.com/corridors/us-27-south-lake)** — constituent places: clermont-florida, leesburg-florida, minneola-florida, groveland-florida ## Published Meeting Readings - [Leesburg Planning Commission — 2025-12](https://zoningsignal.com/meetings/leesburg-pc-2025-12) - [Clermont Planning & Zoning Commission — 2025-10](https://zoningsignal.com/meetings/clermont-pz-2025-10) - [Minneola Planning & Zoning Commission — 2025-09](https://zoningsignal.com/meetings/minneola-pz-2025-09) ## Published Signal Briefs - **[The Self-Storage Canary](https://zoningsignal.com/briefs/self-storage-canary)** — The first commercial-protection move any board makes is on storage — three South Lake boards have now moved _(pattern: The Self-Storage Canary)_ - **[The Quiet Revolution on Highway 27](https://zoningsignal.com/briefs/quiet-revolution-highway-27)** — How four small Florida cities are fighting to control their own futures _(pattern: The Quiet Revolution)_ - **[The Grandfather Window](https://zoningsignal.com/briefs/grandfather-window)** — How a 38-month state regulatory freeze sorted four South Lake cities into protected and exposed _(pattern: The Grandfather Window)_ - **[The Bellwether Gas Station](https://zoningsignal.com/briefs/bellwether-gas-station)** — Two cities, two mechanisms, one converged line on commercial gateway uses _(pattern: The Bellwether Gas Station)_ - **[The Six-Month Board Flip](https://zoningsignal.com/briefs/six-month-board-flip)** — How a single appointment cycle changed Clermont's review philosophy without changing the code _(pattern: The Six-Month Board Flip)_ ## MCP Discovery Server The MCP discovery server lives at `https://zoningsignal.com/.well-known/mcp.json` and exposes eight public, read-only tools: - **describe_zoning_signal()** — the canonical product description, the four artifact types, the regional scope, the methodology. - **list_cities(state?)** — every city with a published place dossier, optionally filtered by state. - **list_corridors()** — every published corridor page. - **current_named_patterns()** — the named patterns currently being tracked across the regional field, sourced from briefs' frontmatter. - **meeting_index(city, from_date?, to_date?)** — meeting readings for a city across an optional date range. - **describe_place(slug, lens?)** — the lens-aware place dossier projection. Lens enum: synthesis (default), developer, investor, broker, attorney, business, resident, civic-leader. - **describe_corridor(slug, lens?)** — the lens-aware corridor projection. Same lens enum. - **semantic_search(q)** — Voyage 4 Lite semantic search across the full published corpus. Returns ranked hits (slug, type, title, similarity, url) using pgvector cosine similarity at threshold 0.55, top 12. Query string maximum 500 characters; accepts GET with `q` param. Tool handlers live at `https://zoningsignal.com/api/mcp/{tool}`. All tools accept GET requests with query parameters; query-shape mirrors the `inputSchema` declared in the discovery manifest. The lens-aware tools are the structural innovation. When an AI agent acts on behalf of a user with a specific cognitive position (*"I'm researching land assembly in south Lake for an investment thesis"*), it can pre-route to `describe_place(slug="clermont-florida", lens="investor")` rather than ingest the full multi-projection surface and synthesize a position from text. The lens projections are crafted artifacts; the agent's position is preserved.