# Agent Trust Firewall (remote · agent-trust-firewall.andrea-vignoni.workers.dev)

Free pre-action policy checks and data cleaning for autonomous AI agents.

- Trust score: 63/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-05

## Components

- remote · `agent-trust-firewall.andrea-vignoni.workers.dev`: 63/100 (this document), [markdown](https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall.md), [page](https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall)

## Channel facts

- Endpoint: `https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp`
- Transports: `streamable-http`
- Auth: `none`
- Version: `1.5.0`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-11-05.

- **Endpoint Security**: 63/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - No authorisation is required to call this server. Nothing on it is marked destructive, but we can't show that an anonymous caller can only read: full credit needs the server to list at least one tool and every tool to be verifiably read-only. We use a best-effort check of tool names and descriptions to tell read-only tools from tools that change things, so some genuinely read-only tools may not get full credit. Names that start with a read verb (get, list, read, search and similar) and accurate readOnlyHint annotations help. Meet that to earn full credit, or require authorisation.
  - HTTPS enforcement could not be verified: the plaintext port answered with HTTP 405, which proves neither a plaintext path nor enforcement.
  - HSTS check failed: the Strict-Transport-Security header is absent.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 61/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 1435 tokens (~287/item across 5 items; 5 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 0/100
  - Stability check failed: schema churn in the 6 days we've observed: 4 tool removals, 0 breaking changes, 0 auth/transport breaks, 5 additions.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 5 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 6 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 60/100
  - Spec-recency check failed: implements MCP spec 2025-06-18; the latest is 2026-07-28.

## Install

### How do I install the Agent Trust Firewall MCP server?

Agent Trust Firewall is a hosted endpoint at https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add --transport http benji18sport-agent-trust-firewall 'https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp'
```

### Cursor

```json
{
  "mcpServers": {
    "benji18sport-agent-trust-firewall": {
      "url": "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp"
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "benji18sport-agent-trust-firewall": {
      "type": "http",
      "url": "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp"
    }
  }
}
```

### Codex

```toml
[mcp_servers.benji18sport-agent-trust-firewall]
url = "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "benji18sport-agent-trust-firewall": {
      "type": "remote",
      "url": "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add benji18sport-agent-trust-firewall --url 'https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  benji18sport-agent-trust-firewall:
    url: "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp"
```

### Netclaw

```json
{
  "McpServers": {
    "benji18sport-agent-trust-firewall": {
      "Transport": "http",
      "Url": "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp"
    }
  }
}
```

### Vellum

```bash
assistant mcp add benji18sport-agent-trust-firewall -t streamable-http -u 'https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp'
```

### Other

```json
{
  "mcpServers": {
    "benji18sport-agent-trust-firewall": {
      "type": "http",
      "url": "https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-11-02 (score 63, 0)

- [functional] Server version: 1.6.1 → 1.8.1

### 2026-11-01 (score 63, +8)

- [security regression] Stability: 0.03 → fail
- [security regression] Tool “data_clean” was removed
- [security regression] Tool “trust_firewall” was removed
- [security regression] Tool “trust_spend_guard” was removed
- [security regression] Tool “trust_validate” was removed
- [security improvement] Authorization: unverified → partial
- [security] The server rewrote its instructions, which are the text every model session reads
- [functional regression] Schema quality: pass → fail
- [functional] First check of Tool coverage: 100
- [functional] First check of Tool coverage: 100
- [functional] Schema quality: fair → excellent
- [functional] Server version: 1.5.0 → 1.6.1
- [functional] New tool “trust_check_action”
- [functional] New tool “trust_check_spend”
- [functional] New tool “trust_clean_data”
- [functional] New tool “trust_get_capabilities”
- [functional] New tool “trust_validate_data”

### 2026-10-31 (score 55, +1)

- [functional improvement] Stability: unverified → 0.03

### 2026-10-30 (score 54)

First indexed and scored.

## MCP tools (5)

### `trust_check_action` (~434 tokens)

Check an action against inline policy

Evaluate a proposed action against evidence, source allowlists, data constraints and payment limits, returning ALLOW/BLOCK with reason codes. All enabled checks must pass: evidence cannot override an over-budget payment. Set action.type to payment to activate spending checks and pass its amount with remaining_budget; contract applies to data, not action. Use this gate before execution, trust_check_spend for budget alone, or trust_validate_data for structure alone. Enforce BLOCK and fail closed on errors. Source labels are unauthenticated; instruction scanning is a limited phrase heuristic. No action execution, wallet access, policy storage or decision history. Free, unauthenticated, stateless pilot: rules apply only to this call. Usage metadata is recorded; request bodies are not stored.

Input parameters:

- `action` (object, required): Proposed action; payment additionally requires amount and remaining_budget.
- `allow_extra` (boolean): With a contract, false rejects top-level data keys not present in the contract. For nested dot paths keep true and enumerate leaf constraints.
- `allowed_merchants` (array): Optional exact, case-sensitive merchant allowlist. An empty list blocks every merchant.
- `contract` (object): Map of dot-separated field paths to constraints, e.g. {"order.total":{"type":"number","min":0,"required":true}}. Inline rules apply only to this request.
- `data` (object): JSON object to inspect; do not submit secrets or personal data.
- `detect_instructions` (boolean): Scan JSON-serialized data for a small fixed phrase heuristic, with possible false positives and negatives.
- `evidence` (array): Caller-supplied evidence labels. Presence is required by default; content and authenticity are not verified.
- `per_transaction_limit` (number): Optional inclusive maximum amount per transaction.
- `remaining_budget` (number): Caller-supplied remaining budget; never read from a wallet or persisted by this service.
- `require_evidence` (boolean): Require at least one evidence item. Setting false intentionally skips the presence check.
- `trusted_sources` (array): Optional exact source allowlist. Every evidence source must be listed.

Output parameters:

- `checks` (array): Names of the check families supported by this gate; optional checks run only when their inputs are present.
- `ok` (boolean): The operation completed; does not imply the verdict is ALLOW.
- `reasons` (array): All failed checks; empty when ALLOW.
- `service` (string): Service that evaluated the request.
- `verdict` (string): Advisory verdict. The integrator must prevent execution on BLOCK or any error.
- `version` (string): Deployed implementation version.
- `warning` (string): Security and enforcement limitations.

### `trust_check_spend` (~266 tokens)

Check a proposed spend against limits

Compare a proposed spend with a remaining budget, transaction ceiling and merchant allowlist, returning ALLOW/BLOCK with reasons. Budget and ceiling are independent: amount=4, remaining_budget=10, per_transaction_limit=3 returns BLOCK; equality passes. Use one currency/unit, and update remaining_budget in your application after each actual purchase: this tool neither tracks cumulative spending nor converts currencies. Omitting allowed_merchants skips merchant checks; supplying it requires an exact merchant match, and [] blocks all merchants. Use trust_check_action to include evidence or data checks. Enforce BLOCK before paying; this tool never accesses a wallet or executes payment. Free, unauthenticated, stateless pilot: rules apply only to this call. Usage metadata is recorded; request bodies are not stored.

Input parameters:

- `allowed_merchants` (array): Optional exact, case-sensitive merchant allowlist. An empty list blocks every merchant.
- `amount` (number, required): Proposed non-negative amount. Use the same currency/unit as both limits.
- `merchant` (string): Merchant identifier; required when allowed_merchants is supplied.
- `per_transaction_limit` (number): Optional inclusive maximum amount per transaction.
- `remaining_budget` (number, required): Caller-supplied remaining budget; never read from a wallet or persisted by this service.

Output parameters:

- `ok` (boolean): The operation completed; does not imply the verdict is ALLOW.
- `reasons` (array): All failed checks; empty when ALLOW.
- `service` (string): Service that evaluated the request.
- `verdict` (string): Advisory verdict. The integrator must prevent execution on BLOCK or any error.
- `version` (string): Deployed implementation version.
- `warning` (string): Advisory-only payment limitation.

### `trust_validate_data` (~262 tokens)

Validate JSON field constraints

Validate JSON field presence, exact types, primitive enums and inclusive numeric bounds, returning ALLOW/BLOCK, field reasons and checked_fields. Combine required:true, type:"integer" and min:1 for a positive mandatory quantity: min alone neither requires a value nor rejects numeric strings. Missing, null or empty optional values skip the other constraints; 0 and false do not. For nested data, fields["order.total"] addresses a leaf; allow_extra:false compares top-level keys, so keep it true for dot-path rules. Use trust_clean_data for transformations or trust_check_action for evidence/spending checks. This validates structure, not facts or provenance. Free, unauthenticated, stateless pilot: rules apply only to this call. Usage metadata is recorded; request bodies are not stored.

Input parameters:

- `allow_extra` (boolean): False rejects top-level keys absent from fields. With nested dot paths use true; this flag is not a recursive additional-properties check.
- `data` (object, required): JSON object to inspect; do not submit secrets or personal data.
- `fields` (object, required): Map of dot-separated field paths to constraints, e.g. {"order.total":{"type":"number","min":0,"required":true}}. Inline rules apply only to this request.

Output parameters:

- `checked_fields` (integer): Number of field rules evaluated.
- `ok` (boolean): The operation completed; does not imply the verdict is ALLOW.
- `reasons` (array): All failed checks; empty when ALLOW.
- `service` (string): Service that evaluated the request.
- `verdict` (string): Advisory verdict. The integrator must prevent execution on BLOCK or any error.
- `version` (string): Deployed implementation version.

### `trust_clean_data` (~315 tokens)

Audit and clean JSON rows

Audit and optionally clean up to 500 JSON rows, returning cleaned_rows, original zero-based issues and counts. Processing order is trim, primary-key deduplication, then required/range checks: keys " a " and "a" collide after trimming. Duplicate removal keeps the first row and skips later-row field checks; missing keys are retained and flagged. REVIEW persists even after repair, so inspect issues before use; [] returns ALLOW. Numeric strings are not converted or range-checked; use trust_validate_data for type enforcement. Only top-level strings are trimmed; inputs are not mutated. Free, unauthenticated, stateless pilot: rules apply only to this call. Usage metadata is recorded; request bodies are not stored.

Input parameters:

- `numeric_ranges` (array): Inclusive bounds for top-level fields that already contain numbers. Use validation for numeric type enforcement.
- `primary_key` (string): Top-level key used for duplicate detection; defaults to id. Keys compare by JSON serialization, so 1 differs from "1".
- `remove_duplicates` (boolean): Drop later rows with the same primary key; missing keys are flagged and retained. Duplicate rows are skipped before required/range checks.
- `required_fields` (array): Top-level field names that must be present and non-null/non-empty. Checked after optional trimming.
- `rows` (array, required): JSON object rows; at most 500. An empty array is valid.
- `trim_strings` (boolean): Trim leading/trailing whitespace from top-level strings; nested values are unchanged.

Output parameters:

- `cleaned_rows` (array): New row objects after the selected transformations.
- `input_count` (integer): Original row count.
- `issues` (array): Audit findings with original zero-based row positions.
- `ok` (boolean): The operation completed; does not imply the verdict is ALLOW.
- `output_count` (integer): Returned row count.
- `service` (string): Service that evaluated the request.
- `verdict` (string): ALLOW: no findings. REVIEW: inspect issues, including removed duplicates. BLOCK: invalid row input.
- `version` (string): Deployed implementation version.
- `warning` (string): Deterministic cleanup limitations.

### `trust_get_capabilities` (~111 tokens)

Discover limits and the stateless workflow

Discover the deployed version, request/row limits, tool names, legacy call aliases and check-clean-recheck workflow before integrating. Pass {}: this inventory endpoint accepts no action, policy or filter arguments; use tools/list for callable schemas and trust_check_action for a verdict. The response explains enforcement and limitations, not stored policy or decision history: rules must accompany each check. Free and unauthenticated, with operational usage metadata recorded but no request-body storage. Re-read after a deployment to detect capability changes.

Output parameters:

- `authentication` (string): No credentials required.
- `limitations` (array): Explicit service limitations.
- `limits` (object): Request limits.
- `mode` (string): Current access mode.
- `ok` (boolean): The operation completed; does not imply the verdict is ALLOW.
- `payments_enabled` (boolean): No billing or payment execution.
- `policy_mode` (string): Supply rules with each call.
- `service` (string): Service that evaluated the request.
- `telemetry` (string): Operational metadata handling.
- `tools` (array): Canonical tools and backward compatible call aliases.
- `version` (string): Deployed implementation version.
- `workflow` (array): Recommended integration steps.

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall#diagnostics

## Score history

- 2026-11-05: 63
- 2026-11-04: 63
- 2026-11-03: 63
- 2026-11-02: 63
- 2026-11-01: 63
- 2026-10-31: 55
- 2026-10-30: 54

## Common questions

### What is the Agent Trust Firewall MCP server?

Agent Trust Firewall is an MCP server listed in the public MCP registry as io.github.Benji18sport/agent-trust-firewall. Free pre-action policy checks and data cleaning for autonomous AI agents. This page covers its hosted endpoint (https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp).

### Is the Agent Trust Firewall MCP server safe to use?

Agent Trust Firewall scores 63 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the Agent Trust Firewall MCP server expose?

Agent Trust Firewall exposes 5 tools: trust_check_action, trust_check_spend, trust_validate_data, trust_clean_data, trust_get_capabilities. Their descriptions and schemas cost roughly 1,388 tokens of context every time the server is loaded.

### Does the Agent Trust Firewall MCP server require authentication?

No. We connected to Agent Trust Firewall without credentials and it answered, so anything it exposes is reachable by anyone who knows the address.

### Is the Agent Trust Firewall MCP server still maintained?

Agent Trust Firewall is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What did VerifyMCP test on the Agent Trust Firewall MCP server?

VerifyMCP ran 12 checks on Agent Trust Firewall, covering TLS certificate, Authorization, HSTS header, DNSSEC and 8 more. One item could not be completed, a gap on our side rather than a finding about Agent Trust Firewall: HTTPS. Supply Chain Security and Provenance & Transparency do not apply to a hosted endpoint, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.

## Links

- Remote endpoint: https://agent-trust-firewall.andrea-vignoni.workers.dev/mcp
- Repository: https://github.com/Benji18sport/agent-commerce-os
- Changelog RSS feed: https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall.xml
- Changelog JSON feed: https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall.json
- HTML version of this page: https://verifymcp.io/servers/benji18sport-agent-trust-firewall/agent-trust-firewall