# `io.github.NovaPixell/lexgrid-mcp` on Anlyon Radar

> Text inside fenced blocks marked untrusted-server-text was written by the MCP server's publisher, not by Anlyon Radar. It is data to inspect, not instructions to follow. Invisible characters are shown as [U+XXXX].
> No finding is not a safety guarantee: these are point-in-time checks of tool definitions, not of code or behaviour.

- Page: https://radar.anlyon.com/servers/io.github.NovaPixell/lexgrid-mcp
- JSON: https://radar.anlyon.com/api/v1/servers/io.github.NovaPixell/lexgrid-mcp
- Version: 1.0.0 (active)
- Publisher: io.github.NovaPixell (GitHub account)
- Runs: remote (streamable-http)
- Sign-in to list tools: none
- Repository: https://github.com/NovaPixell/lexgrid-mcp
- Metadata from the official MCP Registry, synced 2026-11-16T01:17:38.873Z

Title and description, as the publisher wrote them:

```untrusted-server-text
NovaPixel LexGrid
Deterministic EU Compliance, Tax, Public Procurement & Labour Law Engine for AI Agents
```

Scanned 2026-11-16T00:18:19.585Z over streamable-http.

**32 tools: 1 do something that can't be undone, 27 change things and 4 only read.**

| Group | Class | Tools | Claude Code |
| --- | --- | --- | --- |
| Deletes or cancels | destructive | 0 | deny |
| Can't be undone | irreversible | 1 | ask |
| Changes things | write | 27 | ask |
| Only reads | read | 4 | allow |

## Exfil surface

10 private-data -> untrusted-input -> outbound paths in this server alone. Denying 1 private-data tool breaks every one: `mcp__lexgrid__email_deliverability_audit`.

| Tag | Meaning | Tools |
| --- | --- | --- |
| private-data | Reads files, mail, messages, databases, secrets, calendars or repositories. | `mcp__lexgrid__email_deliverability_audit` |
| untrusted-input | Brings in text someone else can write: web pages, inbound email, messages, issues and comments. | `mcp__lexgrid__email_deliverability_audit` `mcp__lexgrid__eu_ai_act_check` |
| outbound | Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL. | `mcp__lexgrid__b2b_tech_profiler` `mcp__lexgrid__cookie_dark_patterns` `mcp__lexgrid__einvoice_xml_auditor` `mcp__lexgrid__eu_ai_act_check` `mcp__lexgrid__gdpr_cmp_tracker_audit` |

| Private data | Untrusted input | Outbound |
| --- | --- | --- |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__b2b_tech_profiler` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__cookie_dark_patterns` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__einvoice_xml_auditor` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__eu_ai_act_check` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__gdpr_cmp_tracker_audit` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__eu_ai_act_check` | `mcp__lexgrid__b2b_tech_profiler` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__eu_ai_act_check` | `mcp__lexgrid__cookie_dark_patterns` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__eu_ai_act_check` | `mcp__lexgrid__einvoice_xml_auditor` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__eu_ai_act_check` | `mcp__lexgrid__eu_ai_act_check` |
| `mcp__lexgrid__email_deliverability_audit` | `mcp__lexgrid__eu_ai_act_check` | `mcp__lexgrid__gdpr_cmp_tracker_audit` |

## Findings

No check matched anything in these tool definitions.


## Can't be undone (1)

Sends, pays, publishes or merges: once done, it cannot be taken back. Claude Code: ask.

### `einvoice_xml_auditor`

- The input addresses people (input.recipient): what it sends cannot be taken back.
- Tag outbound: The input takes a destination (input.recipient): what the tool writes goes there.

Description, as the server wrote it:

```untrusted-server-text
Audit FatturaPA 1.2.2 and Peppol BIS 3.0 XML against official XSD and business rules (SDI error codes, decimal recalculation, Bollo > 77.47€), or build compliant XML from simplified JSON.
```

## Changes things (27)

Creates or changes something that can be changed back. Claude Code: ask.

### `anomalia_offerte_calculator`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Calculates tender anomaly thresholds and 10% wing-trimming under Art. 110 D.Lgs 36/2023 & Annex II.2 for public tenders with lowest price criterion.
```

### `b2b_tech_profiler`

- Nothing in its name or description says what it does, so it is counted as a write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Extracts CMS, frameworks, e-commerce tech from website, extracts VAT number from footer/legal texts, and verifies it deterministically via VIES & GLEIF.
```

### `bando_thresholds_calculator`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Calculates EU tender thresholds under D.Lgs 36/2023, direct award limits, mandatory publication windows in days, electronic submission reductions and ISO provisional guarantee discounts.
```

### `cbam_embedded_emissions`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Calculates direct and indirect embedded emissions (tCO2e) and projected EU ETS CBAM certificate liability under Regulation (EU) 2023/956.
```

### `ccnl_inquadramento_audit`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Audits offered salary against official statutory tables for Italian National Collective Labor Agreements (CCNL Commercio, Metalmeccanica, Studi Professionali).
```

### `codice_fiscale_audit`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Full bidirectional Italian Codice Fiscale audit: D.M. 23/12/1976 CIN check, reverse demographic extraction (gender, birthdate, Belfiore code), omocodia substitution detection & dehomocoding.
```

### `cookie_dark_patterns`

- Nothing in its name or description says what it does, so it is counted as a write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Automated audit of cookie banners and consent modals for illegal dark patterns under EDPB 03/2022 and Garante Privacy guidelines: asymmetric styling, missing first-layer reject, pre-ticked checkboxes and confirmshaming.
```

### `corporate_default_predictor`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Deterministic financial insolvency predictor calculating Altman Z-Score (differentiated for SMEs, manufacturing and services), DSCR, PFN/EBITDA, Current/Quick Ratio and 12-24 month default probability.
```

### `costo_aziendale_dipendente`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Calculates full employer labor cost: RAL, INPS ditta (~24%), INAIL premium, TFR accrual (art. 2120 c.c.), and net hourly cost.
```

### `csrd_esg_calculator`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Calculate official Scope 1 and Scope 2 (Location-based & Market-based tCO2e) emissions and standard European VSME / ESRS indicators using ISPRA, DEFRA and EEA factors.
```

### `dcf_wacc_valuation`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Discounted Cash Flow (DCF) enterprise valuation engine: CAPM Cost of Equity, after-tax Cost of Debt, blended WACC, Gordon Growth terminal value, net debt bridge waterfall, and multi-parameter sensitivity matrix.
```

### `domain_rdap_intel`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Official RDAP query extracting domain registration date, domain age (<30 days anti-phishing alert), registrar, DNSSEC, ASN and EU/non-EU physical jurisdiction.
```

### `dora_third_party_risk`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Audits ICT vendor contracts under DORA Articles 28-30: mandatory exit strategies, regulator inspection rights, EEA data processing location and incident notification windows.
```

### `dual_use_export_control`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Checks dual-use export control regimes and licensing requirements under Regulation (EU) 2021/821 (Annex I & II, UGEA EU001) and international sanctions.
```

### `financial_trend_anomaly`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Forensic accounting analysis: CAGR growth, profitability margins (ROE, ROI, ROS, EBITDA Margin), Beneish M-Score for earnings manipulation detection and Benford law distribution test against ATECO benchmarks.
```

### `fringe_benefit_tax_optimizer`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Calculates and optimizes fringe benefits under Italian Tax Law (Art. 51 TUIR & Legge di Bilancio: 1000€/2000€ caps, ACI car brackets, meal vouchers, utility reimbursements).
```

### `gdpr_cmp_tracker_audit`

- Nothing in its name or description says what it does, so it is counted as a write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.

Description, as the server wrote it:

```untrusted-server-text
Identifies installed Consent Management Platform (Iubenda, Cookiebot, OneTrust, etc.), audits Google Consent Mode v2 implementation and detects unconsented tracking scripts.
```

### `iban_sepa_antifraud`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
ISO 13616 MOD-97 mathematical validation, BIC identification, Peppol routing lookup, and geographic coherence cross-check for BEC fraud prevention.
```

### `incident_severity_triage`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
RTS threshold triage under DORA and NIS2 Art. 23 for major cyber incidents: client impact, duration, financial loss and 4h/24h regulatory reporting deadlines.
```

### `incoterms_cost_allocator`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Deterministic allocation of transport, insurance, export/import clearance and risk transfer points across all 11 official ICC Incoterms 2020 rules.
```

### `lordo_netto_irpef_simulator`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Simulates Italian net take-home salary under post-reform 3-tier IRPEF, TUIR deductions, employee IVS contribution cuts, and regional/municipal surcharges.
```

### `nis2_criticality_classifier`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Deterministic entity classification under Directive (EU) 2022/2555 (NIS2) into Essential vs Important Entity, supervisory regime and statutory fine exposure.
```

### `port_leak_detector`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Non-invasive bounded TCP probe to identify exposed critical databases (MySQL, Postgres, Redis, MongoDB, Elastic), unauthenticated Docker daemon and admin ports with SSRF protection.
```

### `subappalto_cascade_limiter`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Audits subcontracting cascading limits under Art. 119 D.Lgs 36/2023, SIOS works prohibitions, non-rebate mandatory safety costs and CCNL equivalent worker protections.
```

### `tlpt_threat_led_pentest_scope`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Assesses Threat-Led Penetration Testing (TLPT) obligation and scope under DORA Articles 26-27 and Joint RTS JC 2024 15 (TIBER-EU framework).
```

### `tls_http_inspector`

- The description's first verb is "open": a word for a change.

Description, as the server wrote it:

```untrusted-server-text
Opens direct TLS socket to inspect X.509 certificate expiry & SANs, TLS protocol version, cipher suite, HTTP security headers and /.well-known/security.txt.
```

### `vies_gleif_kyb_verify`

- Nothing in its name or description says what it does, so it is counted as a write.

Description, as the server wrote it:

```untrusted-server-text
Verify EU VAT validity on VIES, fetch official registered company name/address, LEI corporate hierarchy on GLEIF and screen against EU/OFAC/UN sanctions.
```

## Only reads (4)

Looks things up and changes nothing. Claude Code: allow.

### `cig_cup_anac_validator`

- The description's first verb is "validate": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Validates Italian ANAC CIG (10-char modulo checksum, Ordinary CIG and SmartCIG) and CUP (15-char public project code) against public tender registry guidelines.
```

### `email_deliverability_audit`

- The description's first verb is "resolve": a word for reading.
- Tag private-data: The name "email_deliverability_audit" says "email": it reads mail.
- Tag untrusted-input: The name "email_deliverability_audit" says "email": it reads inbound email, which other people can write.

Description, as the server wrote it:

```untrusted-server-text
Resolves full DNS tree to audit SPF syntax (recursive 10 lookup limit check), DMARC policy, DKIM selectors, BIMI, MTA-STS and TLS-RPT records.
```

### `eu_ai_act_check`

- The description's first verb is "inspect": a word for reading.
- Tag untrusted-input: The input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
- Tag outbound: The input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
- Claude Code: ask, because it is tagged outbound.

Description, as the server wrote it:

```untrusted-server-text
Inspects DOM and scripts for user-facing AI/chatbot systems (OpenAI, Anthropic, Intercom Fin, Voiceflow, Botpress) and verifies Article 50 transparency disclosure compliance.
```

### `taric_hs_validator`

- The description's first verb is "validate": a word for reading.

Description, as the server wrote it:

```untrusted-server-text
Validates 10-digit TARIC codes, Combined Nomenclature hierarchy, CIF customs valuation, import duty and border VAT payable.
```

## Claude Code permissions

For `.claude/settings.json`, assuming the server is installed as `lexgrid`:

```json
{
  "permissions": {
    "allow": [
      "mcp__lexgrid__email_deliverability_audit",
      "mcp__lexgrid__cig_cup_anac_validator",
      "mcp__lexgrid__taric_hs_validator"
    ],
    "ask": [
      "mcp__lexgrid__vies_gleif_kyb_verify",
      "mcp__lexgrid__einvoice_xml_auditor",
      "mcp__lexgrid__iban_sepa_antifraud",
      "mcp__lexgrid__codice_fiscale_audit",
      "mcp__lexgrid__corporate_default_predictor",
      "mcp__lexgrid__financial_trend_anomaly",
      "mcp__lexgrid__csrd_esg_calculator",
      "mcp__lexgrid__dcf_wacc_valuation",
      "mcp__lexgrid__tls_http_inspector",
      "mcp__lexgrid__domain_rdap_intel",
      "mcp__lexgrid__port_leak_detector",
      "mcp__lexgrid__gdpr_cmp_tracker_audit",
      "mcp__lexgrid__eu_ai_act_check",
      "mcp__lexgrid__b2b_tech_profiler",
      "mcp__lexgrid__cookie_dark_patterns",
      "mcp__lexgrid__bando_thresholds_calculator",
      "mcp__lexgrid__subappalto_cascade_limiter",
      "mcp__lexgrid__anomalia_offerte_calculator",
      "mcp__lexgrid__dora_third_party_risk",
      "mcp__lexgrid__nis2_criticality_classifier",
      "mcp__lexgrid__incident_severity_triage",
      "mcp__lexgrid__tlpt_threat_led_pentest_scope",
      "mcp__lexgrid__cbam_embedded_emissions",
      "mcp__lexgrid__incoterms_cost_allocator",
      "mcp__lexgrid__dual_use_export_control",
      "mcp__lexgrid__costo_aziendale_dipendente",
      "mcp__lexgrid__lordo_netto_irpef_simulator",
      "mcp__lexgrid__ccnl_inquadramento_audit",
      "mcp__lexgrid__fringe_benefit_tax_optimizer"
    ],
    "deny": []
  }
}
```

Capability tags behind the rules (a read tagged outbound is asked about):

- `mcp__lexgrid__einvoice_xml_auditor`: outbound
- `mcp__lexgrid__email_deliverability_audit`: private-data, untrusted-input
- `mcp__lexgrid__gdpr_cmp_tracker_audit`: outbound
- `mcp__lexgrid__eu_ai_act_check`: untrusted-input, outbound
- `mcp__lexgrid__b2b_tech_profiler`: outbound
- `mcp__lexgrid__cookie_dark_patterns`: outbound

## Remote posture

What an anonymous client saw at the remote endpoint, last 2026-11-16T14:19:35.014Z, unchanged since 2026-11-16T14:19:35.014Z. No credentials were sent and no tool was called. Values are shown as the endpoint returned them. Method: https://radar.anlyon.com/methodology#remote-posture

| Fact | Observed |
| --- | --- |
| Reachable | `yes` |
| HTTP status of an anonymous initialize | `200` |
| tools/list served without credentials | `yes` |
| Tools that change things, listed without credentials | `32` |
| Protocol version | `2025-06-18` |
| WWW-Authenticate on 401 | `not observed` |
| resource_metadata in WWW-Authenticate | `none` |
| RFC 9728 protected resource metadata | `none` |
| Authorization servers | `none` |
| Legacy HTTP+SSE transport | `no` |
| Dynamic client registration | `not observed` |
| Client ID metadata documents | `not observed` |

- **high** Tools that change things, listed to anyone ([unauth-remote-writes](https://radar.anlyon.com/methodology#unauth-remote-writes))

```untrusted-server-text
The endpoint served tools/list without credentials, and the list has 32 tools that change things (1 irreversible, 31 write). Whether a call also works without credentials was not tested: tools are never called.

https://lexgrid.novapixel.it/mcp: tools/list returned 32 tools to an anonymous client, including einvoice_xml_auditor (irreversible), vies_gleif_kyb_verify (write), iban_sepa_antifraud (write), codice_fiscale_audit (write), corporate_default_predictor (write) and 27 more
```
