# Inforcer MCP Server

An [MCP](https://modelcontextprotocol.io) server for the [Inforcer](https://www.inforcer.com) public API,
the Microsoft 365 baseline management platform MSPs use to keep client tenants aligned to a
standard. It lets an AI agent read managed tenants, baselines and alignment scores, a tenant's
policies, Secure Score, Entra users / groups / roles and the activity log and, on explicit
request, run assessments and queue reports. Nothing it does changes tenant policies.

- Transport: MCP Streamable HTTP at `POST /mcp`, health probe at `GET /health`
- Stateless and multi-tenant: credentials arrive on every request as HTTP headers and are
  never stored; every tool is listed from the first `tools/list`
- Also provides three MCP prompts: `alignment-review`, `tenant-drift-detail` (tenant),
  `identity-posture-check` (tenant)
- Python 3.12, [uv](https://docs.astral.sh/uv/), FastMCP, httpx
- License: Apache-2.0, Copyright 2026 MSPbots.ai

<div align="center">

## MSPbots Connector

**The MCP Gateway built for MSPs: visible, controlled, and your data stays where it belongs.**

[**Get started at connect.mspbots.ai →**](https://connect.mspbots.ai)

</div>

---

MSPbots Connector gives your AI clients one MCP URL to reach your entire MSP stack (RMM, PSA, documentation, security, finance), and gives you control over who can call which tool and a record of everything that was called.

**Hosted service: [connect.mspbots.ai](https://connect.mspbots.ai)** · Free to use.

### Why Connector

![MSPbots Connector](https://stgoss.mspbots.ai/files/mcp/mspbots-mcp.png)

| | |
|---|---|
| **Visible** | Every tool call is logged: who, which vendor, when. Usage dashboard, CSV export, SIEM streaming. |
| **Controlled** | Role-based, per-tool allowlist. Deny by default. |
| **Protected** | Credentials encrypted per record. AI clients never see your vendor passwords or tokens. |
| **MSP-native** | Built for the MSP stack only: RMM, PSA, documentation, security, finance. |

### Quick start

1. **Sign up** at [connect.mspbots.ai](https://connect.mspbots.ai).
2. **Connect your vendors.** Pick from 150+ connectors and authorize each one on the vendor's own page, or bring your own API keys. No extra vendor accounts needed.
3. **Set access.** Create roles and decide which tools each role may call. Nothing is allowed until you allow it.
4. **Add your MCP URL to your AI client.** Copy the URL from your Connector dashboard into Claude, ChatGPT, Cursor, VS Code, Gemini, Codex, or any MCP-capable client.
5. **Ask your AI to do real work**, and watch every call appear in the audit log.

### Features

**Unified gateway**
- One MCP URL for all your AI clients
- 150+ connectors across the MSP stack
- Custom MCP servers supported, free
- Authorization happens on each vendor's own page; the AI never sees passwords or tokens
- Bring your own credentials: use the vendor API keys you already have
- Multiple accounts per app; connect once, reuse across AI clients
- Tool search: finds the right action for the task
- One sign-in; vendor credentials are injected behind the scenes and OAuth tokens refresh automatically

**Visibility**
- Audit log for every tool call (who, which vendor, when)
- Usage dashboard: call volume, top vendors, error rate, per-user activity
- Export audit logs to CSV, or stream to Loki, Graylog, or LogScale (SIEM)
- Optional recording of tool arguments and result summaries (off by default, enabled per organization)

**Control**
- Per-role, per-tool allowlist, deny by default
- owner / admin / member roles
- SCIM provisioning with Entra ID, Okta, JumpCloud, and Google Workspace

**Security**
- Credentials encrypted one record at a time (AES-256-GCM, per-record key)
- AI clients authenticate with OAuth 2.1 + PKCE
- Vendor responses are scanned for prompt-injection patterns and returned as structured data. This is one layer of defense, not a guarantee against every attack.
- Static egress IP for customer firewall and API allowlisting

### Pricing

**Free.**

Links: [connect.mspbots.ai](https://connect.mspbots.ai) · [mspbots.ai](https://mspbots.ai)

---

## Authentication headers

| Header | Required | Type | Default | Allowed values | Description | Example |
|---|---|---|---|---|---|---|
| `X-Inforcer-Region` | Yes | string | none | `anz`, `eu`, `uk`, `us` (case-insensitive; `api-us.inforcer.com` also accepted) | Region hosting your Inforcer account | `us` |
| `X-Inforcer-Api-Key` | Yes | string | none | - | Inforcer API key | `your-inforcer-api-key` |

Where to get them: an Inforcer API key is issued with scopes. Read-only use needs
`Tenants.Read` (or the narrower `Baselines.Read`, `AlignmentScores.Read`, `tenants.policies.Read`,
`Tenants.Users.Read`, `Tenants.Groups.Read`, `Tenants.Roles.Read`, `Tenants.SecureScores.Read`),
plus `Audit.Read`, `Assessments.Read` and `Reports.Read`; `Assessments.Run` and `Reports.Run` are
needed only for the two run tools.

How the key is sent to Inforcer: every call goes to `https://api-{region}.inforcer.com/api/beta/...`
with the key in the `Inf-Api-Key` header.

A request to `/mcp` without both headers gets HTTP 401 naming the missing one. An unknown region
does not block the request (tools/list still works); every tool call then returns an
`invalid_argument` error naming the allowed regions.

## Tools

All tools return a JSON string. A `tenant` argument accepts the numeric Client Tenant ID, the
Microsoft tenant GUID, the tenant DNS name or its exact friendly name (an ambiguous name returns
the matching ids). **write** marks tools that create work in Inforcer.

| Area | Tools |
|---|---|
| Overview | `inforcer_navigate` (**domain**), `inforcer_status` |
| Tenants | `inforcer_tenants_list`, `inforcer_tenants_get` (**tenant**), `inforcer_tenants_resolve` (**tenant**) |
| Baselines | `inforcer_baselines_list` (baseline_tenant_id) |
| Alignment | `inforcer_alignment_scores`, `inforcer_alignment_details` (**tenant**, baseline) |
| Policies | `inforcer_policies_list` (**tenant**, include_settings) |
| Secure Score | `inforcer_secure_scores_get` (**tenant**) |
| Identity | `inforcer_users_list` (**tenant**, search, continuation_token), `inforcer_users_get` (**tenant**, **user_id**), `inforcer_groups_list` (**tenant**, search, continuation_token), `inforcer_groups_get` (**tenant**, **group_id**), `inforcer_roles_list` (**tenant**) |
| Audit | `inforcer_audit_event_types`, `inforcer_audit_search` (event_types, date_from, date_to, user, page_size, continuation_token) |
| Assessments | `inforcer_assessments_list`; **write**: `inforcer_assessments_run` (**tenant**, **assessment_id**) |
| Reports | `inforcer_reports_types_list`, `inforcer_reports_runs_list`, `inforcer_reports_run_status` (**run_id**), `inforcer_reports_download_output` (**run_id**, **output_id**); **write**: `inforcer_reports_run` (**reports**, **tenants**) |

Errors come back as a JSON envelope instead of an exception:

```json
{"error": {"code": "unauthorized", "message": "Inforcer returned HTTP 403 for beta/tenants/139/policies: Client tenant verification failed", "retryable": false}}
```

`code` is one of `not_configured`, `unauthorized`, `not_found`, `invalid_argument`,
`rate_limited`, `upstream_error`. Inforcer can also answer HTTP 200 with `success: false`; that
is reported as an error too. Reads are retried with backoff on 429, 5xx and network errors;
writes only on 429.

## Run locally

```bash
uv sync
uv run python -m inforcer_mcp          # listens on 0.0.0.0:8080
```

or with Docker:

```bash
docker build --platform linux/amd64 -t inforcer-mcp:dev .
docker run --rm -p 8080:8080 inforcer-mcp:dev
```

Configuration (environment variables, all optional): `MCP_HTTP_HOST` (default `0.0.0.0`),
`MCP_HTTP_PORT` (default `8080`), `LOG_LEVEL` (default `INFO`). Credentials are never read
from the environment.

## Testing example

```bash
# list tools
curl -s http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "X-Inforcer-Region: us" \
  -H "X-Inforcer-Api-Key: your-inforcer-api-key" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'

# alignment of one tenant
curl -s http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "X-Inforcer-Region: us" \
  -H "X-Inforcer-Api-Key: your-inforcer-api-key" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"inforcer_alignment_details","arguments":{"tenant":"example.onmicrosoft.com"}}}'
```

Unit tests (no network access needed):

```bash
uv run pytest
```

## Delivery notes

- **Tool count (24) and prompts (3).** Kept as is because the names are an existing contract.
- **API source.** Inforcer does not publish public API documentation; routes, parameters and
  envelopes follow Inforcer's own OpenAPI document ("inforcer | Public API") as captured by the
  MIT-licensed community PowerShell module InforcerCommunity, and its API reference.
- **Alignment details fixed.** The API requires the baseline to compare against
  (`customBaselineId`); older versions never sent it. `inforcer_alignment_details` now takes an
  optional `baseline` (GUID or exact name) and, when omitted, uses the only baseline the tenant is
  a member of, or lists the candidates.
- **Report run status** reads the run's status from the run list and fetches outputs only for
  finished runs (the outputs route answers 404 while a run is pending, failed or unknown).
- **Report downloads stay small.** csv / html / json outputs are returned as text (first 15,000
  characters); a PDF is returned as base64 only when under 10 KB, otherwise its metadata with a
  note. Older versions inlined any file as base64.
- **Smaller results.** The tenant list leaves out each tenant's policy diff text and licenses
  (`inforcer_tenants_get` has them); the policy list leaves out the full policy settings unless
  `include_settings` is true; baselines report their policy item count instead of the items.
- **New optional parameters:** `baseline` on alignment details, `user` on audit search,
  `include_settings` on the policy list.
- **Tenant names** must match exactly (case-insensitive); an ambiguous name returns the matching
  ids instead of picking one.

## License

Apache License 2.0 - see [LICENSE](LICENSE) and [NOTICE](NOTICE).