> ## Documentation Index
> Fetch the complete documentation index at: https://hc.pillargtm.com/llms.txt
> Use this file to discover all available pages before exploring further.

# MCP Server

> Connect PILLAR to Claude, Cursor, VS Code, and any AI assistant via the Model Context Protocol

# PILLAR MCP Server

PILLAR is the first **Revenue Architecture Operating System (RAOS)** with a native MCP (Model Context Protocol) server. Query your revenue architecture — account health, pipeline, signals, renewals, scoring, plays, tasks, financial cascade, market intelligence, AI-generated narratives, and governed writes — directly from any AI assistant.

## Quick Start

### 1. Generate an API Key

Navigate to **Settings > Integrations** and scroll to the **API Keys & MCP Server** section. Click **Generate API Key** and save the key — it's only shown once.

### 2. Configure Your AI Assistant

Add the following to your Claude Desktop, Cursor, or VS Code MCP configuration:

```json theme={null}
{
  "mcpServers": {
    "pillar": {
      "url": "https://app.pillargtm.com/api/mcp",
      "headers": {
        "Authorization": "Bearer pk_live_your_key_here"
      }
    }
  }
}
```

### 3. Start Querying

Ask your AI assistant natural language questions about your revenue data:

* "What's my pipeline summary?"
* "Which accounts are at risk this quarter?"
* "Show me critical signals and the top 3 save plays that have worked on similar accounts"
* "Simulate what NRR going from 108% to 115% would mean for ARR and AE headcount"
* "What's the top expansion opportunity in the West territory right now?"
* "Ask PILLAR: who's at risk of churning next quarter and why?"

## Endpoint

```
POST https://app.pillargtm.com/api/mcp
GET  https://app.pillargtm.com/api/mcp  (server discovery)
```

Authentication: Bearer API key in the `Authorization` header.

## Available Tools (129)

PILLAR's MCP surface spans **129 tools across 14 categories**, covering every layer of the revenue architecture. The **vertical\_intelligence** category (**63 tools live**) is the competitive moat — it carries PILLAR's canonical district + federal-program datasets that horizontal Revenue AI platforms structurally cannot answer.

<Note>
  **Coverage as of May 2026**: The vertical\_intelligence category is backed by **51 jurisdictions** (50 states + DC + federal) and **26 federal datasets** (8 IPEDS components + 8 Higher Ed sources + 10 K-12 sources), exposed through **63 MCP tools**. Per-district coverage is at 51/51 jurisdictions for assessment proficiency (5.03M cells across \~19,700 LEAs), cohort graduation (391k cells), accountability status (24k cells), and engagement/chronic absenteeism (104k cells) — the four priority district-grain tables. **K-12 state funding allocations (NEW): 46 of 51 jurisdictions (90.2%), 114,699 per-LEA rows, $957B+ captured from primary state aid programs (CA LCFF, TX FSP, NY Foundation Aid, IL EBF, FL Net State FEFP, OH Net State Funding, WA BEA, MI Bulletin 1014, etc.).** **HE state aid (NEW): 7,876 per-institution rows across 58 jurisdictions, $7.94B captured — IPEDS SFA per-institution state grant aid (3,669 institutions FY22-23 + 3,693 FY21-22) plus 11 state-specific programs (TX TEXAS Grant, CA Strong Workforce, 9 MI scholarship/grant programs).** Per-state DOE deep ingest covers 27+ states at the recent-year grain; federal EDFacts SY 2020-21 backfill closes the long tail to 51/51 for the priority surfaces. The schema, ingest pipeline, MCP wrappers, and 550+ build-time-enforced Guarantees are runtime-truth — every commit blocks merge unless the canonical-shape validators (`G-X-31` through `G-X-40`) accept every row landing in the 26 federal-data + 47 state-funding tables. All Round 8 tools (Scorecard Field-of-Study, FSA CDR/GE/HCM/distress-score, SHEEO SHEF, NC-SARA, Carnegie 2025, IPEDS HR/ADM/AY/AL/EF-CIP, CCD School Universe, EDGE Locale, CRDC ×2, OSEP IDEA-B, McKinney-Vento, Title III, Migrant Ed, Perkins V, NSLP-CEP, NIEER) are live and MCP-callable today.
</Note>

The list below is representative — the source of truth is the tool catalog at `src/lib/mcp/tool-catalog.ts`, and new tools ship continuously as vertical-intelligence surfaces (K-12 state-calendar procurement windows, federal Title program eligibility, cooperative-contract lookups, NCES district enrichment, state DOE assessment + accountability + graduation) come online.

### Tier 0 — Core Surface

The foundational tool set that shipped with the original MCP server. Still the most-invoked tier for day-to-day agent workflows.

#### Core Intelligence

| Tool                   | Description                                                                                                         |
| ---------------------- | ------------------------------------------------------------------------------------------------------------------- |
| `get_dashboard`        | Full GTM health snapshot: ARR, pipeline, NRR, at-risk accounts, signal count, forecast health                       |
| `get_pipeline_summary` | Pipeline totals: open, commit, best-case, weighted pipeline, NRR, ARR                                               |
| `search_accounts`      | Search accounts by name, segment, or territory with health/risk/priority scores                                     |
| `get_account_360`      | Full account detail: scores, contacts, signals, opportunities, district intelligence                                |
| `get_active_signals`   | Active signals filtered by severity (CRITICAL/WARNING/INFO) or family (RENEWAL/PIPELINE/EXPANSION/ACCOUNT/COVERAGE) |
| `get_account_health`   | Health, risk, and priority scores with scoring decomposition showing which factors drive each score                 |
| `get_renewal_risk`     | Upcoming renewals with risk scores, filtered by days out (30/60/90)                                                 |

#### Operational

| Tool                      | Description                                                                                                        |
| ------------------------- | ------------------------------------------------------------------------------------------------------------------ |
| `get_forecast`            | Quarterly forecast: won, commit, probable, upside with override amounts and rep-level detail                       |
| `get_revenue_bowtie`      | Full acquire → close → retain/expand funnel with conversion rates at each stage                                    |
| `get_territory_economics` | Territory P\&L: revenue, costs, yield ratios, health classification per territory                                  |
| `get_leads`               | Lead funnel with ICP fit scores, behavioral scores, and funnel stage. Filter by status or score threshold          |
| `acknowledge_signal`      | Update a signal's status (ACKNOWLEDGED, IN\_PROGRESS, RESOLVED). *Write.*                                          |
| `create_play`             | Create an intervention play for an account (SAVE\_PLAY, EXPANSION\_PLAY, ONBOARDING\_PLAY). *Write.*               |
| `complete_play`           | Complete a play with mandatory outcome (RENEWED, CHURNED, EXPANDED, LOST) and auto-update linked renewal. *Write.* |
| `get_board_report`        | Executive summary for board meetings: ARR, NRR, GRR, pipeline coverage, risk distribution, EBITDA                  |
| `get_scoring_rules`       | Scoring rule summary with weights and categories                                                                   |

#### Data Readiness

| Tool                   | Description                                                                                                                                                                  |
| ---------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `get_data_readiness`   | Check if CRM data is ready for scoring: readiness level (ready/degraded/blocked), scoring-ready account count, blocking issues                                               |
| `get_affected_records` | Identify specific accounts with data quality issues. Filter by issue code (no\_contacts, no\_recent\_activity, missing\_segment, zero\_arr, score\_imbalance, negative\_arr) |

#### Flywheel & Benchmarks

| Tool                          | Description                                                                                                                        |
| ----------------------------- | ---------------------------------------------------------------------------------------------------------------------------------- |
| `get_calibration_stats`       | Flywheel scoring calibration dashboard: outcome distribution, weight drift from baseline, top rules by calibration impact          |
| `get_calibration_history`     | Scoring weight calibration history: pending feedback count, current weights, and chronological calibration event log               |
| `get_benchmark_percentiles`   | Blueprint benchmark percentile rankings: pillar scores and operational metrics ranked against anonymized peer cohort               |
| `list_benchmark_cohorts`      | List available benchmark cohorts with member counts and snapshot data. Cohorts group orgs by industry, revenue range, and maturity |
| `get_benchmark_opt_in_status` | Check whether the organization is opted in to anonymized benchmarking and which cohort is assigned                                 |

#### Connector Observability

| Tool                       | Description                                                                                                                                          |
| -------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- |
| `list_connectors`          | List all configured data connectors across CRM, email/calendar, product usage, and support ticketing with per-provider status, last sync, and health |
| `get_sync_log`             | Recent connector sync runs from the unified sync log. Shows successes, failures, records synced, and timing. Filter by connector name                |
| `get_account_data_sources` | Per-account breakdown of which data sources contributed to its score: CRM, product analytics, support tickets, or contracts provenance               |

### Tier A — Plays, Tasks, Expansion (8 tools)

#### Plays Intelligence

| Tool                       | Description                                                                                                                           |
| -------------------------- | ------------------------------------------------------------------------------------------------------------------------------------- |
| `list_plays`               | List plays in flight. Filter by account, state (PENDING, ACTIVE, COMPLETED, DISMISSED), or owner. Answers "what's already in flight?" |
| `get_play_effectiveness`   | Template-level win rate, ARR impact, health delta, and which templates are underperforming. Filterable by segment/tier/territory      |
| `get_play_recommendations` | System-suggested plays per account based on signal patterns, scoring, and historical effectiveness                                    |

#### Tasks & Activities

| Tool                    | Description                                                                                                             |
| ----------------------- | ----------------------------------------------------------------------------------------------------------------------- |
| `list_tasks`            | Platform task list with filters (assignee, status, account, source, priority). Includes summary counts and overdue flag |
| `get_activity_timeline` | Email, call, meeting, task, and play event timeline. Filter by account, user, or date range                             |

#### Expansion Whitespace

| Tool                     | Description                                                                                                                   |
| ------------------------ | ----------------------------------------------------------------------------------------------------------------------------- |
| `get_expansion_summary`  | Org-wide expansion whitespace: total addressable expansion ARR, top candidate accounts, product-level whitespace distribution |
| `get_expansion_signals`  | Layer-I signals tagged as expansion triggers (adoption surge, champion promoted, new use case detected)                       |
| `get_expansion_affinity` | Product co-adoption matrix and cross-sell probability per product pair                                                        |

### Tier B — Financial Cascade (9 tools)

| Tool                       | Description                                                                                                                                            |
| -------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ |
| `simulate_nrr_impact`      | Simulate ARR / headcount / ROI impact of moving NRR from current to target. Returns AE and CSM-equivalent headcount, retain-vs-acquire cost comparison |
| `get_revenue_impact`       | ARR-at-stake analysis across renewal window, expansion opportunity value, churn exposure, net revenue position                                         |
| `get_budget_variance`      | GTM budget variance vs plan per period (rep cost, marketing spend, CS cost)                                                                            |
| `get_procurement_calendar` | Procurement windows by account: fiscal year alignment, renewal overlap. Filterable by territory/segment/upcoming                                       |
| `get_procurement_forecast` | Procurement-window-aware forecast (different from stage-based `get_forecast`)                                                                          |
| `get_cohort_curves`        | Customer retention + NRR cohort curves by acquisition period. Investor-grade                                                                           |
| `list_cohort_definitions`  | Available cohort slicing definitions (acquisition quarter, segment, deal size, territory)                                                              |
| `get_forecasted_renewals`  | ML-forecasted renewal outcomes combining base rates + current risk scores                                                                              |
| `get_customer_readiness`   | Implementation and adoption readiness: onboarded, stuck, milestone status, time-to-value risk                                                          |

### Tier C — Market Intelligence (5 tools)

| Tool                        | Description                                                                                               |
| --------------------------- | --------------------------------------------------------------------------------------------------------- |
| `get_tam_sam`               | TAM / SAM / SOM sizing by vertical, filterable by fiscal year                                             |
| `get_territory_equity`      | Territory fairness analysis: ARR balance, account distribution, quota-to-pipeline ratios                  |
| `simulate_headcount_change` | Project ARR uplift, ramp lag, cost impact, and breakeven quarter of adding/removing reps                  |
| `get_district_intelligence` | K-12 district enrichment (enrollment, demographics, bond measures, superintendent changes) via Starbridge |
| `get_data_health`           | Org-wide data health dashboard: freshness lag, orphan counts, duplicate counts, pipeline blockers         |

### Tier D — AI Orchestration (7 tools)

| Tool                            | Description                                                                                                   |
| ------------------------------- | ------------------------------------------------------------------------------------------------------------- |
| `ask_pillar`                    | Ask PILLAR's native RAG layer a question. Returns grounded answer with citations from CRM + signals + scoring |
| `list_narratives`               | List cached AI-generated narratives (board briefings, forecast commentary, account deep-dives)                |
| `get_narrative`                 | Full prose body + citations for a specific narrative                                                          |
| `generate_action_plan`          | AI-authored per-account action plan: owner assignments, sequenced tasks, expected outcomes, SLAs              |
| `generate_account_intelligence` | Multi-source account brief (CRM + signals + scoring + district intel + contact web)                           |
| `generate_coaching_insights`    | Rep-level forecast accuracy, deal-velocity patterns, stuck-deal flags, 1:1 talking points                     |
| `generate_board_narrative`      | Board-grade prose from the current dashboard data                                                             |

### Tier E — Scoring Transparency (5 tools)

| Tool                     | Description                                                                                                               |
| ------------------------ | ------------------------------------------------------------------------------------------------------------------------- |
| `get_scoring_backtest`   | Historical accuracy of scoring models: how well did risk/priority/expansion scores predict actual outcomes over 12 months |
| `get_scoring_profile`    | Org's customized scoring rule weights and active configuration                                                            |
| `get_forecast_weights`   | Custom forecast probability weights per stage + category mappings                                                         |
| `get_contracts_config`   | Contracts object configuration: date models, term lengths, expansion vs new-biz classification                            |
| `get_scoring_thresholds` | Org's custom score-severity thresholds (CRITICAL vs WARNING vs INFO per model)                                            |

### Tier F — Governed Writes (5 tools)

| Tool                      | Description                                                                                                                            |
| ------------------------- | -------------------------------------------------------------------------------------------------------------------------------------- |
| `create_task`             | Create a platform task. Source type distinguishes NBA / signal / play / MQL handoff / manual. Governance log entry automatic. *Write.* |
| `update_task`             | Update status, assignee, priority, or due date. *Write.*                                                                               |
| `update_signal_status`    | Set signal to ACKNOWLEDGED / IN\_PROGRESS / RESOLVED / SNOOZED / DISMISSED with optional reason. *Write.*                              |
| `set_renewal_disposition` | Set renewal disposition directly (RENEWED / CHURNED / EXPANDED / LOST / PENDING). *Write.*                                             |
| `opt_in_benchmarks`       | Opt the org in or out of anonymized peer-cohort benchmarking. *Write.*                                                                 |

### Tier G — Vertical Intelligence (63 tools live)

<Info>
  **The competitive moat.** Horizontal revenue platforms (Salesforce, HubSpot, Gong, Clari) cannot ship these tools because they don't maintain federal education datasets, accreditor calendars, per-state procurement timing, or canonicalized state DOE assessment data. PILLAR does — the Tier G surface answers questions the customer's own procurement office and CFO operate in.
</Info>

K-12 + HigherEd specific tools backed by maintained federal datasets (IPEDS — now including the 5 IPEDS-extension components Human Resources / Admissions / Academic Year Tuition / Academic Libraries / Enrollment by CIP — plus NCES CCD, EDFacts, F-33 district finance, College Scorecard institution + field-of-study, Carnegie Classification 2025, FSA Cohort Default Rates / Gainful Employment / NSLDS aggregate / HCM, SHEEO SHEF, NC-SARA, CRDC, OSEP IDEA Part B, McKinney-Vento, Title III ELA, Migrant Education, Perkins V CTE, NCES EDGE Locale Codes, NSLP/CEP, NIEER State of Preschool), accreditor cycles, state fiscal calendars, Title program formulas, and **all 50 states + DC canonicalized into a unified per-LEA assessment / accountability / graduation surface (51/51 jurisdictions covered for assessment proficiency, cohort graduation, accountability status, and engagement/chronic absenteeism via per-state DOE deep ingest for 27+ states + federal EDFacts SY 2020-21 backfill closing the long tail)**.

**K-12 — State DOE academic outcomes (NCES LEAID-keyed, 51 jurisdictions unified)**

| Tool                                       | Description                                                                                                                                                                                                                                                                                                                             |
| ------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `get_district_assessment_proficiency`      | Per-LEA proficiency cell (subject × grade × subgroup × year) across all 51 jurisdictions (27+ states via per-state DOE deep ingest at recent-year grain; federal EDFacts SY 2020-21 fills the remaining 24 jurisdictions to 51/51), normalized to a unified `pct_proficient_or_above` column with documented per-state policy footprint |
| `get_district_assessment_trend`            | Year-over-year proficiency trend with `continuity_break` flags across assessment-family transitions (PARCC→MCAP, FSA→FAST, AIR→Cambium)                                                                                                                                                                                                 |
| `get_district_assessment_benchmarks`       | District proficiency vs state aggregates (P25 / median / P75) AND NAEP state + national, with mandatory `standards_comparability_note` whenever cross-family figures juxtapose                                                                                                                                                          |
| `get_district_assessment_subgroup_gaps`    | Ranked gap-from-all-students per subgroup; suppression envelope preserved (suppressed cells appear with `gap_pct: null`, never silently dropped)                                                                                                                                                                                        |
| `get_district_accountability_status`       | Canonical CSI / TSI / LRAP / REC / GS status with the `is_intervention_priority` boolean derived from a single source of truth                                                                                                                                                                                                          |
| `get_district_graduation_metrics`          | Multi-year-window graduation rates (4yr / 5yr / 6yr / 7yr) with explicit `extension_coverage` documenting which extensions are available                                                                                                                                                                                                |
| `get_district_engagement_metrics`          | Chronic absenteeism, attendance, dropout per LEA × subgroup with worst-subgroup ranking                                                                                                                                                                                                                                                 |
| `get_district_advanced_coursework_metrics` | AP/IB/dual-credit enrollment + pass rates with `ZERO_PASSING_NOTE` flag for the "access without success" pattern                                                                                                                                                                                                                        |
| `get_district_ccmr_metrics`                | College/Career/Military Readiness composites with explicit `CCMR_COMPOSITE_NOTE` warning that composites aren't comparable across states                                                                                                                                                                                                |
| `get_district_growth_metrics`              | YoY improvement scores with `GROWTH_VS_LEVEL_NOTE` to prevent conflating growth with absolute proficiency                                                                                                                                                                                                                               |
| `get_district_early_childhood_metrics`     | Pre-K + early literacy with `coverage_scope_note` (frequently public-school-enrolled only)                                                                                                                                                                                                                                              |
| `get_district_graduation_pathway_metrics`  | TX-specific graduation pathway codes with state-attribution preservation (no silent cross-state collapse)                                                                                                                                                                                                                               |
| `get_district_refusal_rates`               | NY 3-8 assessment opt-out rates per LEA × subject × subgroup                                                                                                                                                                                                                                                                            |
| `get_state_naep_comparison`                | State assessment trend vs NAEP trend for the 11 Tier-1 states; surfaces `naep_disagrees: true` when state cut-score recalibrations diverge from federal benchmark                                                                                                                                                                       |
| `get_ingestion_health`                     | Per-state ingest freshness + last-publish lag                                                                                                                                                                                                                                                                                           |

**HigherEd — IPEDS institutional intelligence (UNITID-keyed)**

| Tool                                   | Description                                                                                                                                                                                                            |
| -------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `get_ipeds_enrollment_trend`           | 10-year enrollment trend for a higher-ed institution (by IPEDS UNITID): total + FTE headcount, Pell-eligible share, first-generation share, trend direction, CAGR, and a computed enrollment-cliff risk score (0-100). |
| `score_institution_tuition_dependency` | Tuition-and-fees / total-revenue ratio over 5 years with dependency score. Higher = more exposed to enrollment-cliff → vendor ability-to-pay risk.                                                                     |
| `get_pell_grant_institutional_share`   | % of undergrads receiving Pell, total Pell dollars, 5-year trend, and federal-budget-sensitivity score. Higher Pell reliance = more exposure to federal appropriation risk.                                            |

**K-12 — Federal Title dollars (NCES LEAID-keyed)**

| Tool                             | Description                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |
| -------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `get_district_title_allocations` | Per-district federal Title dollars. Title I-A nationally via F-33 C14 (\~15.6k districts × 5 FYs); Title III-A via F-33 C36 (\~5.6k districts × 5 FYs); aggregate `OTHER-FED-RESTR` bucket via F-33 C25 (\~15k districts × 5 FYs — bundles II-A / IV-A / V / VI / VII / VIII / CTE non-Perkins / ESSER); Titles II-A and IV-A as a 3-district sample (LAUSD / Chicago / NYC) from EDFacts. Every row carries `source`, `reporting_lag_months`, and the aggregate bucket has `is_aggregate: true` with `aggregate_includes` listed so DRAFTER can't mistake it for a specific Title. |

**Public-sector — Procurement calendars + compliance**

| Tool                                     | Description                                                                                                                                                                                                             |
| ---------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `get_fiscal_year_procurement_windows`    | When a given state's K-12 districts open and close their buying windows — and which months are off-limits for procurement conversations.                                                                                |
| `get_state_higher_ed_budget_cycle`       | A state's HigherEd budget calendar: legislative session windows, governor sign-off month, earliest disbursement month.                                                                                                  |
| `get_federal_title_programs`             | Catalog of ESEA / ESSA Title programs (I-A, I-B, I-C, I-D, II-A, III-A, IV-A, IV-B, V, VI, VII, VIII) with FY2025 federal appropriations, ESSA evidence-tier requirements, and product-category fit for EdTech vendors. |
| `get_accreditation_review_cycle`         | A higher-ed institution's next accreditation self-study, site visit, and decision dates across 8 regional + national accreditors (SACSCOC, HLC, MSCHE, NECHE, NWCCU, WSCUC, ACCJC, DEAC).                               |
| `score_cooperative_contract_eligibility` | Whether a target district / institution can buy through a named cooperative (NASPO ValuePoint, TIPS, Sourcewell, CRC, OMNIA Partners, E\&I, MHEC, PEPPM, etc.) — turns a 3-month procurement cycle into 30 days.        |

#### The canonicalization claim

> PILLAR canonicalizes **51 jurisdictions** (50 state DOEs + DC + federal) with a documented policy footprint, a structural honesty layer, and **sixteen independent layers of accuracy verification**:
>
> **Round 1-5 reconciliation layer (closes "is the state-DOE proficiency number right?")**
>
> * **Macro-level reconciliation** against state-published statewide aggregates with **24-state coverage** (`G-X-25`)
> * **Micro-level spot-checks** against **17 hand-validated district fixtures across 13 states** including the load-bearing LDOE R36→036 alias (`G-X-26`)
> * **External NAEP trend-direction cross-validation** for the 11 Tier-1 states with a live MCP route at `/api/vertical/state-naep-comparison` (`G-X-27`)
> * **Silent-corruption canary** on every ingested cell with queue-backed weekly review via the `value_unknown_alarms` table (`G-X-28`)
> * **Federal Title pass-through reconciliation** between EDFacts allocations and SEA-published disbursements (`G-X-29`)
> * **Per-district Title allocation spot-checks** closing the loop on "we know proficiency AND federal allocation are right for the same district" (`G-X-30`)
>
> **Round 8 federal-data canonical-shape layer (closes "is the federal-dataset row right?")**
>
> * **IPEDS-extension shape discipline** for Human Resources / Admissions / Academic Year Tuition / Academic Libraries / Enrollment by CIP — including biennial-even-year discipline on EF-CIP and pre-2014 collection\_status discipline on Academic Libraries (`G-X-31`)
> * **OPEID padding integrity** on the institution\_crosswalk join — the only authorized path between UNITID-keyed (IPEDS, Scorecard, Carnegie) and OPEID-keyed (FSA CDR/GE/NSLDS/HCM, NC-SARA) datasets (`G-X-32`)
> * **College Scorecard shape** validators on institution-level + field-of-study tables (`G-X-33`)
> * **Carnegie 2025 four-dimension derivation discipline** — `is_r1`/`is_r2` MUST be derivable from `research_activity_designation`; SAEC eligibility flag MUST be true when SAEC classification is present (`G-X-34`)
> * **FSA regulatory-status discipline** preventing accidental publish-rate inference during the 2019-2023 GE rescission gap; CDR status enum + HCM level enum locked (`G-X-35`)
> * **SHEEO SHEF + NC-SARA** state-level shape with USPS-keyed JSONB integrity (`G-X-36`)
> * **CRDC biennial discipline** — collection year MUST be even; suspensions ≤ 2× total enrollment sanity check (`G-X-37`)
> * **CCD School Universe + EDGE locale enum** — title\_i\_status, charter\_status, magnet\_status, virtual\_indicator, locale\_code all locked to documented value sets (`G-X-38`)
> * **OSEP IDEA Part B + K-12 federal program** state-aggregate shape (`G-X-39`)
> * **NCES EDGE entity-type-conditional ID-length** (school=12 / lea=7-10 / postsecondary=1-6 digits) + **NIEER 0-10 quality benchmark hard cap** (`G-X-40`)

Each verification layer is a build-time-enforced Guarantee with a cited spec entry — see [The Guarantee → Vertical Intelligence (X)](/the-guarantee#vertical-intelligence-x-the-canonicalization-layer) for the full structural enforcement chain.

**Runtime-truth status (April 2026)**: 550 Guarantee tests pass on every commit; 0 route type errors; 26 federal datasets ingested with 890,000+ canonical rows ready for live upsert across 35 ingest scripts.

#### Why this differentiates PILLAR

State DOEs each express proficiency on a different scale, suppression with different sentinels, accountability in 4-tier vs 5-tier vs A-F, with subgroup labels that vary across all 51 jurisdictions. Each state DOE essentially publishes data that's only legible inside its own bureaucracy. Without the canonicalization layer, a query like *"show me districts with declining ELA proficiency under 50%, chronic absenteeism above 20%, and accountability rating in the bottom two tiers"* is structurally impossible across state lines — every comparison would require state-specific knowledge of cut-scores, sentinels, and subgroup mappings. PILLAR's vertical intelligence MCP layer makes that query a single tool call that runs in milliseconds and returns the unified answer with the comparability caveats baked in.

**Provenance + honest coverage limits.** Every Tier G response carries row-level `source` strings and a response-level `data_provenance` section documenting reporting lag and known gaps. For example, `get_district_title_allocations` discloses the three dead-end paths that prevent national per-LEA II-A / IV-A coverage (the Q1 2026 ed.gov site reorg deleted the per-state workbooks; Internet Archive has zero XLSX snapshots of the ed.gov Title paths; USAspending.gov records only state-level primary awards and double-counts carryover obligations). Similarly, the per-state assessment ingestion documents per-state policy footprints (TN "Approached/Met/Exceeded" vs LA "Mastery and above" vs WI "Advanced+Meeting") and surfaces `continuity_break` flags whenever year-over-year comparison crosses an assessment-family transition. Consumers should never guess at data availability — the route tells them what it has and what it doesn't.

## Compatible AI Assistants

PILLAR's MCP server works with any MCP-compatible client:

* **Claude Desktop** — Anthropic's AI assistant
* **Claude Code** — CLI development assistant
* **Cursor** — AI-first code editor
* **VS Code** — GitHub Copilot with MCP support
* **ChatGPT Desktop** — OpenAI's assistant
* **Windsurf** — Codeium's AI editor
* **Zed** — High-performance editor with MCP

## Security

* API keys are stored as SHA-256 hashes — the raw key is never persisted
* Each key is scoped to a single organization via Row-Level Security
* Keys can be revoked instantly from Settings > Integrations
* All data is org-scoped — you can only query your own organization's data
* MCP server inherits PILLAR's multi-tenant isolation
* All tools are defined in a single source-of-truth catalog at `src/lib/mcp/tool-catalog.ts` which both the external MCP server and the in-app Drafter runtime consume. No drift between surfaces.

## Example Conversations

**CRO Morning Briefing:**

> "What does my GTM health look like this morning?"
>
> Claude calls `get_dashboard` → returns ARR, pipeline, at-risk accounts, signal count, forecast health

**Renewal Review with NRR simulation:**

> "Which renewals are at risk in the next 30 days — and what would saving all of them do to NRR?"
>
> Claude calls `get_renewal_risk` with `days_out: 30` → totals at-risk ARR → calls `simulate_nrr_impact` with current and projected NRR to show ARR uplift + AE-equivalent headcount

**Account Deep Dive with Effectiveness:**

> "Tell me about Houston ISD and recommend the best save play template for a district like this"
>
> Claude calls `search_accounts` → `get_account_360` → `get_play_effectiveness` filtered to the account's segment/tier → recommends the highest-win-rate template with evidence

**Signal Triage with Task Creation:**

> "Show me critical signals, acknowledge the top one, and create a follow-up task for the account owner"
>
> Claude calls `get_active_signals` with `severity: "CRITICAL"` → `update_signal_status` → `create_task` with source\_type: "signal"

**Procurement-aware Forecast:**

> "What's our forecast by procurement window instead of CRM stage?"
>
> Claude calls `get_procurement_forecast` → explains the difference vs stage-based forecast → flags which deals are likely to slip based on procurement alignment

**Board Briefing Generation:**

> "Generate a board-grade narrative for Tuesday's meeting and pull the cohort retention curves to go with it"
>
> Claude calls `generate_board_narrative` → `get_cohort_curves` → composes the pack

**Scoring Credibility Check:**

> "How accurate has PILLAR's renewal risk score been historically?"
>
> Claude calls `get_scoring_backtest` → reports hit rate, false positives, and calibration drift

**Connector Health Check:**

> "Are all my data sources syncing correctly?"
>
> Claude calls `list_connectors` → `get_sync_log` for any connectors showing errors

**Score Provenance:**

> "Where does the health score for Houston ISD come from?"
>
> Claude calls `get_account_data_sources` → returns which values came from CRM vs product analytics vs support tickets

## Rate Limits

The MCP server inherits PILLAR's standard API rate limits. For most tools, this is effectively unlimited for normal AI assistant usage patterns. If you encounter rate limiting, contact [support](/support).