# fxray (pypi · fxray)

Power Apps via pac: solutions, canvas Power Fx source, Dataverse metadata, FetchXML, gated imports.

- Trust score: 68/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-09-26

## Components

- pypi · `fxray`: 68/100 (this document), [markdown](https://verifymcp.io/servers/s-curvelabs-fxray/fxray.md), [page](https://verifymcp.io/servers/s-curvelabs-fxray/fxray)

## Channel facts

- Registry: `pypi`
- Package: `fxray`
- Version: `0.1.0`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, 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 2026-09-26.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - Runs hatchling.build at install time, a recognised native-build step with no shell scripting around it.
  - 1 of 30 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 2 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 74/100
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 2705 tokens (~117/item across 23 items; 23 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 not yet verified: not enough scan history yet (needs a 30-day window).
- **Tool Coverage**: 67/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 0% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.
  - An AI judge read all 24 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a current MCP spec version (2026-07-28).

**Unverified: 1 category.** A category scored 0 because we could not verify it: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.

## Install

### How do I install the fxray MCP server?

fxray runs locally as a PyPI package, launched with uvx fxray. 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 s-curvelabs-fxray -- uvx fxray
```

### Cursor

```json
{
  "mcpServers": {
    "s-curvelabs-fxray": {
      "command": "uvx",
      "args": [
        "fxray"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "s-curvelabs-fxray": {
      "command": "uvx",
      "args": [
        "fxray"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add s-curvelabs-fxray -- uvx fxray
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "s-curvelabs-fxray": {
      "type": "local",
      "command": [
        "uvx",
        "fxray"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add s-curvelabs-fxray --command uvx --arg fxray
```

### Hermes

```yaml
mcp_servers:
  s-curvelabs-fxray:
    command: "uvx"
    args: ["fxray"]
```

### Netclaw

```json
{
  "McpServers": {
    "s-curvelabs-fxray": {
      "Transport": "stdio",
      "Command": "uvx",
      "Arguments": [
        "fxray"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add s-curvelabs-fxray -t stdio -c uvx -a fxray
```

### Other

```json
{
  "mcpServers": {
    "s-curvelabs-fxray": {
      "command": "uvx",
      "args": [
        "fxray"
      ]
    }
  }
}
```

## 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.

### 2026-09-25 (score 68, +3)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-09-24 (score 65, +15)

- [security improvement] Malware scan: unverified → pass

### 2026-09-23 (score 50)

First indexed and scored.

## MCP tools (23)

### `pac_status` (~49 tokens)

Check the setup: config (allowlisted environments, writes enabled, output folder), the pac CLI version, and
pac's auth profiles. Run this first when anything fails. Does not contact an environment.

### `whoami` (~62 tokens)

Who pac is signed in as and which org it reaches for an allowlisted environment (`pac env who`).
Fails with instructions if the sign-in expired. `environment` = URL or id from the allowlist (default: first).

Input parameters:

- `environment` (string)

### `list_environments` (~41 tokens)

All Dataverse environments the signed-in account can see (global discovery), each marked allowed / not allowed.
Only allowed ones can be used by the other tools.

### `list_solutions` (~83 tokens)

Solutions in an allowlisted environment: unique name, display name, version, managed flag, publisher.
Managed (mostly Microsoft system) solutions are hidden unless include_managed=true. name_filter is a
case-insensitive substring on unique or display name.

Input parameters:

- `environment` (string)
- `include_managed` (boolean)
- `name_filter` (string)

### `solution_components` (~119 tokens)

What a solution contains.
Offline (source= a solution zip or unpacked folder): manifest (name, version, managed, publisher/prefix), root
components by name, canvas apps, flows, connection references, environment variables, missing dependencies.
Live (solution= unique name): every component row from the environment with type and, for tables, canvas apps,
flows, environment variables and connection references, the resolved name (others show the object id).

Input parameters:

- `environment` (string)
- `solution` (string)
- `source` (string)

### `list_canvas_apps` (~65 tokens)

Canvas apps: live in an environment (`pac canvas list`: name, created by, modified on), or offline in a
solution zip / unpacked folder (logical name, display name, AppVersion stamp).

Input parameters:

- `environment` (string)
- `source` (string)

### `export_solution` (~121 tokens)

Export a solution from an allowlisted environment (read-side: nothing changes in the environment) into a new
timestamped folder under the output dir, and by default unpack it there (pac solution unpack) so its tables,
relationships and canvas apps can be read offline. managed=true exports the managed package (only possible from
the environment where the solution is unmanaged). Takes ~1 minute. Returns the zip and folder paths.

Input parameters:

- `environment` (string)
- `managed` (boolean)
- `solution` (string, required)
- `unpack` (boolean)

### `unpack_solution` (~70 tokens)

Unpack a solution zip (pac solution unpack, local only) into a new folder under the output dir
(or <output>/unpacked/<folder_name>). Returns the folder, which works as source= for the offline tools.

Input parameters:

- `folder_name` (string)
- `zip_path` (string, required)

### `solution_check` (~91 tokens)

Run the Power Apps solution checker on a solution zip (pac solution check): uploads the zip to Microsoft's
checker service (a cloud call, nothing changes in the environment), waits, and summarizes the SARIF results by
rule and severity. Takes several minutes. Output is kept under <output>/checks/.

Input parameters:

- `environment` (string)
- `geo` (string)
- `zip_path` (string, required)

### `diff_solutions` (~114 tokens)

What changed between two versions of a solution (each a zip or unpacked folder, e.g. two export_solution
outputs): version/publisher/root components, tables and columns (type, required level, length, choices),
relationships, canvas apps (screens, controls, formula changes as small diffs, data sources), flows/env vars/
connection references, and changed files for anything else (views, forms).

Input parameters:

- `max_formula_changes` (integer)
- `new` (string, required)
- `old` (string, required)

### `canvas_overview` (~90 tokens)

Summary of a canvas app's source: screens (in Studio order) with control counts, components, control types,
data sources and app checker counts. source = .msapp, SourceCode folder (Src/*.pa.yaml), or a solution zip /
unpacked folder (then app= logical or display name when it holds several apps).

Input parameters:

- `app` (string)
- `source` (string, required)

### `list_controls` (~90 tokens)

The control tree (name, type, variant, parent, depth, number of properties), optionally for one screen and/or
one control type (substring, e.g. 'Gallery', 'Button'). Use get_control for formulas.

Input parameters:

- `app` (string)
- `control_type` (string)
- `limit` (integer)
- `screen` (string)
- `source` (string, required)

### `get_control` (~78 tokens)

One control (or screen, or 'App') with its type, variant, parent, children and property formulas (Power Fx).
properties= limits to named properties (e.g. ['OnSelect', 'Items']).

Input parameters:

- `app` (string)
- `control` (string, required)
- `properties`
- `source` (string, required)

### `search_formulas` (~139 tokens)

Search every Power Fx formula in a canvas app for a substring (or regex=true). code_only=true ignores matches
inside string literals and comments. property= restricts to one property name (e.g. 'OnSelect'); control= to
controls whose name contains it. Returns control, property, and the matching lines.

Input parameters:

- `app` (string)
- `case_sensitive` (boolean)
- `code_only` (boolean)
- `control` (string)
- `limit` (integer)
- `pattern` (string, required)
- `property` (string)
- `regex` (boolean)
- `source` (string, required)

### `canvas_data_sources` (~111 tokens)

The app's data sources (Dataverse tables, connectors, static data) with the formulas that reference each one.
Flags UNREFERENCED data sources, which Power Apps Studio silently drops on the next save/publish. On a solution
source it also checks meta.xml DatabaseReferences (empty = every write breaks after import) and reports columns
the solution defines but the app's cached table schema lacks (invisible to the published player).

Input parameters:

- `app` (string)
- `source` (string, required)

### `app_checker_results` (~84 tokens)

The App Checker results Power Apps Studio saved inside the app (AppCheckerResult.sarif): counts by level and
issues grouped by rule with locations (screen.control.property). level= filters (Critical/High/Medium/Low).

Input parameters:

- `app` (string)
- `level` (string)
- `limit` (integer)
- `source` (string, required)

### `list_tables` (~134 tokens)

Dataverse tables. With source= (solution zip/folder, or canvas .msapp/SourceCode folder for the tables the app
caches): name, display name, entity set, column counts. Without source: live table NAMES from the environment
(the entity catalog via FetchXML; pac has no live column metadata, so describe_table needs a source).
pattern = substring on logical or display name.

Input parameters:

- `app` (string)
- `custom_only` (boolean)
- `environment` (string)
- `limit` (integer)
- `pattern` (string)
- `source` (string)

### `describe_table` (~137 tokens)

Columns of one table: logical name, type, display name, required level, max length, lookup targets, choice
values; plus primary id/name columns and relationships. `table` = logical, display or entity-set name.
System columns (createdby, versionnumber, ...) are hidden unless include_system_columns=true. full=true adds
descriptions, formats and numeric ranges. source: solution zip/folder, or a canvas app (its cached schema).

Input parameters:

- `app` (string)
- `full` (boolean)
- `include_system_columns` (boolean)
- `source` (string, required)
- `table` (string, required)

### `list_relationships` (~85 tokens)

Table relationships from a solution (or a canvas app's cached schema): one-to-many (referenced/one side,
referencing/many side and its lookup column, cascade delete) and many-to-many (intersect table).
table= limits to relationships touching that table.

Input parameters:

- `app` (string)
- `source` (string, required)
- `table` (string)

### `fetch_query` (~143 tokens)

Run a read-only FetchXML query against an allowlisted environment (pac env fetch) and return rows as dicts.
FetchXML cannot modify data. Rows are capped at max_rows (default from config, ceiling 5000): pac itself ignores
top/count/page and would page through every record, so the server stops reading at the cap. pac also ignores
\<order>; the server sorts client-side when the result is complete. Values are pac's formatted strings. Prefer
filters and explicit <attribute> lists; the primary key column is always appended.

Input parameters:

- `environment` (string)
- `fetchxml` (string, required)
- `max_rows` (integer)

### `pack_solution` (~116 tokens)

Pack an unpacked solution folder into a solution zip (pac solution pack). Local only: nothing is sent to an
environment. dry_run=true (default) returns the plan and command plus warnings (canvas data sources Studio would
prune, empty DatabaseReferences). Default zip goes under <output>/packs/. Executing needs allow_writes=true.

Input parameters:

- `dry_run` (boolean)
- `folder` (string, required)
- `overwrite` (boolean)
- `package_type` (string)
- `zip_path` (string)

### `import_solution` (~203 tokens)

Import a solution zip into an ALLOWLISTED environment. dry_run=true (default) returns the plan: target org
(verified with pac env who), package name/version/managed, what is installed now, whether this is a new install,
an update or an upgrade, blockers and warnings (managed/unmanaged mismatch, downgrade, canvas AppVersion trap,
empty DatabaseReferences), backup and the exact pac command. mode: 'update' (default; never deletes), 'upgrade'
(managed only; DELETES components missing from the zip), 'holding'. Executing (dry_run=false, allow_writes=true)
first exports the installed unmanaged solution as a backup and returns an undo recipe.

Input parameters:

- `dry_run` (boolean)
- `environment` (string)
- `force_overwrite` (boolean)
- `mode` (string)
- `publish_changes` (boolean)
- `settings_file` (string)
- `zip_path` (string, required)

### `publish_customizations` (~63 tokens)

Publish all customizations in an allowlisted environment (pac solution publish). dry_run=true (default)
returns the verified target and command. Executing needs allow_writes=true and dry_run=false.

Input parameters:

- `dry_run` (boolean)
- `environment` (string)

## Diagnostics

Captured diagnostic sections: Provenance, Install scripts, Dependencies. The full working is on the page: https://verifymcp.io/servers/s-curvelabs-fxray/fxray#diagnostics

## Score history

- 2026-09-26: 68
- 2026-09-25: 68
- 2026-09-24: 65
- 2026-09-23: 50

## Common questions

### What is the fxray MCP server?

fxray is an MCP server listed in the public MCP registry as io.github.S-CurveLabs/fxray. Power Apps via pac: solutions, canvas Power Fx source, Dataverse metadata, FetchXML, gated imports. This page covers its PyPI package (fxray).

### Is the fxray MCP server safe to use?

fxray scores 68 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 26 September 2026. 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 fxray MCP server expose?

fxray exposes 23 tools: pac_status, whoami, list_environments, list_solutions, solution_components, and 18 more. Their descriptions and schemas cost roughly 2,288 tokens of context every time the server is loaded.

### Is the fxray MCP server still maintained?

fxray is still listed as active in the MCP registry. We last reached this channel on 26 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the fxray MCP server under?

fxray declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- PyPI project: https://pypi.org/project/fxray/
- Socket report: https://socket.dev/pypi/package/fxray
- Repository: https://github.com/S-CurveLabs/fxray
- Changelog RSS feed: https://verifymcp.io/servers/s-curvelabs-fxray/fxray.xml
- Changelog JSON feed: https://verifymcp.io/servers/s-curvelabs-fxray/fxray.json
- HTML version of this page: https://verifymcp.io/servers/s-curvelabs-fxray/fxray