Mirror · Overview
Mirror Essence
A measurement system for AI-driven discovery. Brand Discovery Intelligence™ in concise form.
The Discovery Shift
For two decades, brand discovery has been measured through traditional search. Keyword rankings. Traffic. Share of voice. Those metrics describe one system.
Today, three systems generate answers when a consumer asks a question:
- Search engines — Google, Bing
- Answer engines — AI Overviews, voice assistants
- Generative AI — ChatGPT, Claude, Perplexity
Consumers do not distinguish between them. They ask a question and expect an answer. Brands continue to measure only one. The result is a growing gap between brand strength and brand visibility at the moment decisions are formed.
The Mirror Method
Mirror measures brand discoverability across all three systems.
- AEO · Answer Engine Optimization · 35% — Eligibility to be extracted as a direct answer to consumer questions.
- GEO · Generative Engine Optimization · 35% — Presence in AI-generated responses describing a brand.
- SEO · Search Engine Optimization · 30% — Visibility within traditional search results.
These three engines are scored independently, then combined into a single composite — the AI Citability Score (ACS), a weighted measure of current brand discoverability — and read through the AI Citability Gap (ACG), a diagnostic view that shows where the largest performance gap exists.
Mirror does not measure traffic. Mirror measures whether a brand is present when answers are given.
Measurement Discipline
- Determinism — engine scores computed in code; the model only confirms signals
- Model Calibration — Claude Sonnet 4.6 · Temperature 0
- Rubric Versioning — v3.0 · August 2026
- Structural Integrity — Three engines, every layer
Mirror is the first instrument within the category of Brand Discovery Intelligence™.
daniels ai design studio · Stowe, Vermont · v3.0 · August 2026
Mirror · Working Environment
Studio
Behind the Looking Glass — Mirror's working environment. Six steps, six named actors, one continuous flow.
Foreword
Mirror reveals what is present and what is missing in the answers consumers receive about a brand at a moment in time. Studio produces and places the content that improves discoverability in the engines, and Shadow tracks how brand authority moves over time.
Studio runs a six-step production flow — Findings, Plan, brand.com, Copywriter, Media, Make. Each step has a named actor. The brand and Daniels work together. AI handles the production load. Humans direct what gets generated, what ships, and where the work goes live.
Studio is built on a discipline. The reasoning is visible. The work is auditable. Authority stays with the brand at every consequential pause.
The Workflow
Studio runs a continuum of six steps. Each step has a defined actor and a defined role.
Discover
Mirror examines how the brand is expressed across search engines, answer engines, and generative AI. The audit produces findings, not directions.
Measure
Findings are scored against the ACS Method rubric. Reproducibility targets define what counts as variance and what counts as material signal.
Create
The brand and Daniels work together through Studio to author the response. Themes are selected. Page plans, copy, and media architectures are drafted with AI handling the production load.
Generate
Studio produces the artifacts — page plans by theme, content by type, channel allocations grounded in the audit's findings. Generation happens on explicit human direction. Nothing auto-fires.
Production
The brand decides what ships and where. Daniels supports placement and execution. The work goes live in the world under the brand's authority.
Track
Subsequent audits land in Shadow. Score progressions tell whether the work moved the brand. The loop closes here, and the next priority surfaces.
Findings to Make
Studio runs a defined production flow that begins with the audit's findings and ends with the work moving to placement. Six steps, each with a named actor.
Findings
The audit's observations, presented for the brand and Daniels to read together. The widest gap leads. Findings are not directions; they are what Mirror revealed.
Plan
Mirror's reading of which gaps matter most, organized by engine and ordered by priority focus. The brand decides which gaps to address first.
brand.com
Page architectures generated theme by theme. Each selected theme produces a complete answer-first page plan with deployment recommendations. Plans accumulate so the brand can compare and choose.
Copywriter
Content generated across formats — FAQ, blog articles, social posts, email copy, ad copy — drawn from the audit's findings. Each format is generated on explicit selection. Nothing auto-fires.
Media
Channel allocation against the brand's budget, with architecture grounded in the audit's diagnosis rather than agency convention.
Make
The work moves to placement. The brand decides what ships and where. Daniels supports execution. The work goes live in the world under the brand's authority.
Six steps, six named actors, one continuous flow. Studio is the room.
How the Brand and Daniels Work Through Studio
Studio is open. The reasoning is visible. The work is auditable.
A Daniels engagement is not a black box. The brand sees what Studio sees. The audit findings are visible. The page plans are visible. The copy is visible. The media architecture is visible. The Shadow audit history is visible. There is no surface where Daniels acts without the brand being able to read what is happening.
AI handles the production load that humans used to coordinate by hand. Page plans for each theme. Content across formats. Media allocations grounded in the audit. The work that previously required coordination across separate vendors happens inside Studio, with the brand and Daniels directing what gets produced.
The brand's authority is preserved at every consequential pause. Theme selection is the brand's. Generation is initiated by explicit click, not by the system. What ships is the brand's decision. Daniels supports — surfaces what the audit found, helps the brand work through the priorities, produces the work the brand directs. The brand decides.
Signal Discipline
Studio operates inside a governed perimeter. Five rules define how outputs are produced and how authority is preserved.
- Signals must be authorized. Studio's reasoning grounds in the audit pipeline. Findings are not invented; they are read from what Mirror's instrument captured.
- Generation happens on explicit direction. Nothing auto-fires. Page plans are generated when a theme is selected. The pause between intent and generation is the brand's decision moment.
- Plans accumulate; nothing overwrites. When the brand selects a second theme, the first theme's plan does not disappear. The workspace lets the brand hold multiple plans side by side and choose.
- The reasoning is visible. Every output traces back to the findings that produced it. The brand can read why a page plan is structured a particular way.
- Authority flows through the brand at every consequential pause. Daniels supports. AI generates. Studio surfaces. The brand decides what ships.
The Continuum Holds
Mirror reveals. Studio supports the work. Shadow tracks whether the work moved the brand. The continuum runs from discovery through tracked outcome, with the brand's authority preserved at every consequential pause.
Mirror is an audit instrument. Studio is the working environment that surrounds it. Together they form the continuum that defines Brand Discovery Intelligence™ in practice.
Studio · Working Environment · v3.0
Mirror · Measurement Integrity
Shadow
Score Progression and the Audit-to-Outcome Loop. Mirror's measurement-integrity layer.
Foreword
Mirror reveals what is present and what is missing in the answers consumers receive about a brand at a moment in time. Studio produces and places the content that improves discoverability in the engines, and Shadow tracks how brand authority moves over time.
Movement without measurement discipline is noise. Movement with measurement discipline is signal. Shadow holds the difference.
Shadow pairs with Studio. Studio is where work gets created. Shadow is where work gets verified. Together they form the back half of the Mirror continuum: discover, measure, create, generate, production, and track.
What Shadow Records
Every audit Mirror runs lands in Shadow. Each record carries the full context of the measurement.
- Engine scores — AEO, GEO, and SEO scored independently on a 0–100 scale, each anchored to the rubric in force at the time of the audit.
- Capture timestamp — An ISO timestamp marking when the audit was run. Millisecond-precise, drives Shadow's chronological ordering.
- Rubric version — The version of the Mirror Measurement Spec under which the audit was scored.
- Calibration parameters — The model and temperature used for the audit. Currently Claude Sonnet 4.6 at temperature 0.
- Brand context — The brand's industry, primary domain, and the audit's findings — the qualitative observations that produced the scores.
The audit corpus is Mirror's institutional memory. Every record is traceable; nothing is fabricated; manual fabricated readings are not accepted into Shadow because the methodology's reproducibility claim depends on every signal being authorized by the audit pipeline.
Variance versus Material Signal
Mirror does not publish a fixed ± band. Under rubric v3.0 the engine score is computed in code from confirmed signals, so the arithmetic does not vary at all. What varies is how much could be observed on a given run — and that is reported openly as the score range.
A two-point change between consecutive audits is not a story. A move from Average to Above Average is.
The Audit-to-Outcome Loop
Shadow is where the loop closes between Mirror's measurement instrument and the work that follows from it.
- Audit lands in Shadow. Mirror runs an audit. The findings, scores, and full audit context land in Shadow.
- Brand and Daniels work through Studio. Plan organizes findings by engine and priority. The brand decides which gaps to address first.
- Work goes live. The brand decides what ships and where. The work goes live in the world under the brand's authority.
- Subsequent audits land in Shadow. The new scores land alongside the prior audits. The score progression becomes visible.
- Movement is interpreted against the rubric. Within-band movement is variance. Band crossings are material signal. Shadow flags only what the rubric says is real.
- The next priority surfaces. The new audit's findings show what's still present, what's still missing, and where attention should focus next.
Mirror remains accountable to its own next audit. The brand does not have to take Mirror's word for whether the work moved the brand — Shadow shows it.
Why Shadow Exists
Shadow is what separates a measurement system from a one-time analysis.
A one-time analysis produces a snapshot. The reader receives a score, an interpretation, and a set of recommendations. The analysis is judged on its argument, not on its outcomes. A measurement system holds itself accountable. Every audit lands in the same place. Every score is comparable to the one before it. Every movement is interpreted against published reproducibility targets.
Shadow is Mirror's accountability surface. Three commitments live here:
- Audits are versioned to the rubric in force. When the rubric changes, the version increments and is documented. Past audits are not re-scored against new rubrics.
- Reproducibility targets are published. Per-engine variance is named in advance. The brand knows what counts as noise before any audit runs.
- Material signal is rule-defined, not interpretation-defined. The brand reads movement the same way Daniels does.
The Methodology Defends Itself
Mirror is built to be a reference standard for Brand Discovery Intelligence™. A reference standard earns its name through discipline that holds across time, not through the elegance of a single audit.
Shadow holds the discipline. The audit corpus accumulates. The reproducibility targets stay published. The Interpretation Rule holds. Every brand engagement adds to the institutional memory that makes Mirror a measurement system rather than a moment of analysis.
Shadow · Measurement-integrity layer · v3.0
Yellow Belt ai · Operating Discipline
Yellow Belt ai for Mirror
Running the Brand Discovery Intelligence™ continuum under SIGS discipline. The operator training for running Mirror's six-beat continuum on a brand — Reflection through Track — with governance holding at every step.
What this is
Yellow Belt ai is the operator training discipline for Mirror and Brand Discovery Intelligence™. Yellow Belt ai for Mirror teaches a person to run Mirror's continuum on a brand with discipline, governance, and reproducibility.
It is not a new framework. It is the operating discipline for running the system the studio has already named and defined: Reflection reads the brand, Studio creates the work that addresses what the audit revealed, Shadow tracks whether the work moved discoverability over time.
The loop
Mirror's continuum runs in six beats. Each maps to a published Studio step and a named actor. Nothing here is invented — the loop is the canon, learned as practice. Governance is not one of the beats. SIGS and QC run across all six — the discipline that lets the loop be trusted, and one day, partly automated.
The six beats
Reflection (Discover). Mirror reads the brand across AEO, GEO, and SEO and returns a Reflection — findings, not directions. Every Reflection is run fresh, reasoned through from scratch.
Understand the gaps (Measure). The findings are scored against the ACS Method rubric. Each engine carries a Performance Gap of 100 minus its score. Movement within a band is variance; movement across a band is material signal.
North Star (AI Citability Gap → Plan). The largest Performance Gap is the priority focus — the single target you steer by. "North Star" is the Yellow Belt word for the canon's priority focus.
Create the files (Create + Generate). Through Studio's brand.com, Copywriter, and Media actors, the selected theme becomes answer-first page plans, copy by format, and media allocation.
Publish (Production / Make). The brand decides what ships and where. The work goes live under the brand's authority.
Track (Shadow). The next Reflection lands in Shadow alongside the prior ones. Score progression shows whether the work moved the brand, read against the publish history.
Governance across every beat — SIGS and QC
SIGS — the Signal Integrity & Governance Specification — holds five disciplines that apply at every beat:
- Authorization before interpretation
- Domain containment
- Intelligence before action
- Human authority preservation
- Bounded automation
No finding is invented; it is read from what the instrument captured. No work fires without explicit human direction. QC runs alongside — measure, log, audit, override, version, monitor drift.
This is where a bad finding is caught before it ships: an override at the human-authority pause before Production, under SIGS and QC — not a function of Shadow. Shadow is Track only.
Worked example — a hospitality brand
One turn of the loop, run by hand.
Reflection. Mirror read the brand and returned AEO 38, GEO 42, SEO 52 — an AI Citability Score (ACS) of 44 (Average).
Understand the gaps. The Performance Gaps: AEO [62], GEO [58], SEO [48].
North Star. The widest gap is AEO [62], so the AI Citability Gap names AEO the priority focus. The Plan leads with Answer-First Content Architecture.
Create the files. Studio produced answer-first FAQ Q&A and FAQPage schema for the brand's core lodging and amenity pages — replacing a FAQ page left stale from a prior season.
Governance. The Reflection's draft answer included a superlative — a "only [category] in [region]" claim — that, on verification, did not survive scrutiny. A peer property in the same region held a comparable claim. The override applied at the human-authority pause before Production: the brand's permitted superlative was narrowed to a verifiable, location-specific framing. A bad finding caught by SIGS and QC, before it shipped — not by Shadow.
Publish. The brand places the corrected work live under its own authority.
Track. The next Reflection lands in Shadow. If AEO crosses from Below Average toward Average, the rubric calls it material signal — proof the work moved the brand, and the next turn of the loop can begin.
Passing Yellow Belt
You hold the belt when you can run the full continuum on a brand — Reflection through Track — with SIGS discipline holding at every beat: findings grounded in the instrument, no invented claims, human authority preserved before Production, drift monitored by QC. The belt is earned by practice, not by reading.
Yellow Belt ai for Mirror · v2 · June 2026
Mirror · Agent Ecosystem
Mirror in the Agent Ecosystem
How Mirror — the studio's brand-discovery instrument — operates where brands are now discovered: by AI agents, through the Model Context Protocol.
Mirror is the MCP for Brand Discovery Intelligence™.
Foreword
The web's readers changed. Brands used to be found by people. Increasingly they are found by agents — software that asks, retrieves, and answers on a person's behalf. Mirror measures whether AI cites a brand. To do that work where the work now happens, Mirror runs as an MCP server: an instrument other agents can call directly.
The instrument that measures AI-citability must itself be AI-citable.
Why we built an MCP server
The unit of brand value moved. It was once the composition — the image, the page, the story, made legible to people. It is now the entity — structured, cited, machine-legible, made legible to the systems that mediate attention.
Mirror practices what it measures: it is now a resolved entity in the Knowledge Graph — a Wikidata item wired to Daniels AI Design Studio and to The ACI 55. The instrument that measures citability is itself citable.
If agents discover and judge brands, the studio that measures that discoverability must be reachable by agents — not only by people. The MCP server is Mirror's hands in the agent layer. It is also a posture: be found, not chase. The studio publishes an instrument and lets the agents come to it.
Mirror in the registry
Mirror is published in the Model Context Protocol registry — the place agents look for tools. Four sit on the surface. score is open and anonymous — a Lite reading: a grade band and an estimated range, the free check that wins discovery and builds footprint. reflect is the full diagnosis — the complete Brand Reflection — gated and metered, and the only level that returns a precise, citable AI Citability Score (ACS). aci55 looks up The ACI 55 — the studio's published index of leading brands ranked by AI Citability Score — for any brand's standing in it. mcp_engine measures the fourth engine itself: whether a brand is callable by agents, not merely readable. Lite estimates. Full measures.
The plan is not traffic. The plan is to become the instrument agents reach for by default when a brand's citability is in question. Footprint becomes adoption; adoption becomes the standard.
Own the measurement, and you own the category.
New tools, deliberately
The surface will grow — under one rule: self-serve the diagnosis, provide the answers. Tools that expand measurement may join the registry: tracking a brand's citability over time; consulting the studio's intelligence directly. The cure — building the fixes, distributing them, watching the score move — stays human-side, where judgment lives. Every tool ships metered and bounded. The studio adds capability without surrendering control.
SIGS in the design
Mirror's MCP server is governed by SIGS — the studio's discipline for delegating to machines without losing the wheel. Three principles, written into the build:
- Authorization before interpretation — a tool that spends or diagnoses runs only behind an authenticated, consented connection. The channel authorizes before the engine acts.
- Bounded automation — a hard daily spend ceiling, per-caller rate limits, and a fixed free allowance gate every call before it reaches the engine. The meter is enforced at the door, not recorded after.
- Human authority preservation — a person approves the connection; the answers are provided by people, not automated; nothing publishes without the studio's word.
Governance is not a setting. It is a feature.
Brand Discovery Intelligence™ agents
A class is forming. As agents proliferate, a category emerges whose job is to make brands discoverable and citable to other AIs — Brand Discovery Intelligence™, delivered by agents.
Daniels AI builds two. Beckett, the studio's named intelligence. And a Brand Discovery Agent, built as a controlled instrument — a known quantity against which the rest of the ecosystem can be measured. The studio does not only measure citability. It is defining the agent class that delivers it.
Daniels AI designs. Beckett understands. Humans decide.
Visit the Mirror MCP server →
Mirror · Agent Ecosystem · mcp.danielsdesignstudio.com/mcp · Rubric v3.0 · 2026
CoveBud · Overview
CoveBud Essence
CoveBud is the Botanical Understanding Dashboard — an AI-native intelligence surface for cannabis consumers, operators, and farms.
What CoveBud Is
One of Daniels AI Design Studio's live product systems. CoveBud connects live dispensary data, strain intelligence, consumer discovery, and operator insight into one governed intelligence platform. The system operates against real Vermont dispensary menus, not synthetic data.
CoveBud is not a forecast. CoveBud is an operating system surface — synced, normalized, and answering from what is actually on the shelf.
Read the full document on CoveBud →
CoveBud Essence · 2026
CoveBud · Architecture
Architecture
CoveBud's production stack is a multi-model agentic pipeline that ingests, normalizes, and reasons over live cannabis data from dispensaries across Vermont. Three working capabilities.
The Three Capabilities
CoveBud Connect — a headless connector mesh polling dispensary menus in real time. Strain Entity Resolution — fuzzy matching that resolves messy product names to canonical strain identities. CoveBud AI Chat — retrieval-augmented generation grounded in hyperlocal data.
Three layers, one pipeline. CoveBud answers from real synced data, not generic guesses.
Read the full document on CoveBud →
CoveBud Architecture · 2026
CoveBud · Future State
Autonomous Cultivation
Self-custody compute, air-gapped inference, zero chemical inputs. CoveBud's terminal architecture is a fully autonomous cultivation system.
The Stance
Every environmental variable — sensed, modeled, actuated by on-premise intelligence. Zero outbound connection. The compute layer runs on local self-custody hardware: inference-optimized edge accelerators executing quantized open-weight models, fine-tuned on cultivar-specific grow data. No cloud dependency.
The farm's intelligence lives on the farm's hardware, owned by the farmer.
Read the full document on CoveBud →
Autonomous Cultivation · Future State · 2026
CoveBud · Strategy
Why Vermont
CoveBud launched in Vermont because Vermont is small enough to know completely, and grows cannabis good enough to make the system worth building.
The Smart Choice
Two reasons. Scope — one regulatory boundary, a knowable retail footprint, a consumer base that fits. The constraint is the product. Reputation — Vermont has earned its standing as one of the country's top cannabis-growing states. Small cultivators, named genetics, craft knowledge that exists because the cannabis here is grown by people who care which strain it is.
CoveBud is in Vermont because Vermont was the smart choice. Hyperlocal by design, not by default.
Read the full document on CoveBud →
Why Vermont · Strategy · 2026