ARIA Customer Experience MCP Server
PyPI distribution: aria-cx-mcp ·
Hosted endpoint: https://mcp.aria.simplification.io/mcp ·
Marketing page: aria.simplification.io/mcp
Expose ARIA intelligence as Model Context Protocol tools consumable by Claude Desktop, Cursor, and any MCP-compatible client.
v1.9.5 highlight: 6 packaged Playbooks (/aria:morning-brief,
/aria:churn-watch, /aria:agency-rollup, /aria:dsar-assist,
/aria:bench-review, /aria:compliance-snapshot) ship as MCP prompts/*.
See PLAYBOOKS.md for the full reference.
Naming note (v1.9.5 W5): the PyPI distribution name is
aria-cx-mcp(renamed fromaria-mcp-serverto dodge a permanent collision with an unrelated clinical-research package on PyPI). The internal Python package name (aria_mcp_server) is unchanged. The new console script isaria-cx-mcp; the legacyaria-mcp-serverscript is retained for one minor release before removal in v1.9.6.
Connect to Claude Desktop — hosted (recommended)
Paste into claude_desktop_config.json:
{
"mcpServers": {
"aria": {
"url": "https://mcp.aria.simplification.io/mcp",
"transport": "streamable-http",
"headers": {
"Authorization": "Bearer aria_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
}
}
}
}
The hosted endpoint runs on Railway with a Cloudflare-managed
mcp.aria.simplification.io domain. TLS, audit-log emission, and tier
gating are enforced server-side.
Connect to Claude Desktop — local install (advanced)
For offline / air-gapped use:
pip install aria-cx-mcp
aria-cx-mcp --configure
Then add to claude_desktop_config.json:
{
"mcpServers": {
"aria": {
"command": "aria-cx-mcp"
}
}
}
Or set environment variables:
export ARIA_API_URL=https://api.aria.simplification.io
export ARIA_API_KEY=aria_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Run a hosted instance (Docker)
docker build -t aria-mcp .
docker run -p 8787:8787 \
-e ARIA_API_URL=https://api.aria.simplification.io \
-e ARIA_API_KEY=aria_live_xxxxx \
aria-mcp
The container exposes /mcp (streamable-HTTP transport) and /health.
Tools (21)
Source of truth: aria_mcp_server/tools/__init__.py TOOL_SCHEMAS.
| # | Tool | Description |
|---|---|---|
| 01 | aria_search_customers |
Search customers by name / email / external id |
| 02 | aria_get_conversations |
List recent conversations for a customer |
| 03 | aria_get_memory |
Retrieve customer memory record |
| 04 | aria_classify_message |
Run the 12-step pipeline on a new message |
| 05 | aria_search_kb |
Semantic KB search |
| 06 | aria_get_analytics |
Analytics (volume, retention, LTV, flow perf) |
| 07 | aria_invoke_agent |
Invoke an agent by ID with arbitrary input |
| 08 | aria_create_flow |
Generate a draft flow from NL |
| 09 | aria_get_suggestions |
Fetch proactive suggestions |
| 10 | aria_get_outcomes |
Fetch bench agent outcomes |
| 11 | aria_refine_flow |
Refine an existing flow with NL |
| 12 | aria_configure |
Unified Copilot — configure anything via NL |
| 13 | aria_get_agent_frameworks |
Framework badges + EU AI Act Article 13 transparency |
| 14 | aria_explain_decision |
Per-decision explainability (powers the "Why?" button) |
| 15 | aria_get_grounding |
Grounding citations + peer-review record |
| 16 | aria_list_playbooks |
List marketplace Playbooks with filters |
| 17 | aria_submit_playbook |
Submit a Playbook YAML for marketplace review |
| 18 | aria_get_watch_status |
Proactive Watch health grid (7 dimensions) |
| 19 | aria_list_open_alerts |
Paginated open watch alerts |
| 20 | aria_recommend_starter_kit |
Recommend best-fit Starter Kit |
| 21 | aria_get_attribution |
Attribution rollups (last-touch / linear / time-decay) |
Playbooks (6)
Source of truth: aria_mcp_server/playbooks/__init__.py PLAYBOOK_DEFINITIONS.
| # | Playbook | Tier | Description |
|---|---|---|---|
| 1 | aria:morning-brief |
Starter+ | Daily attention brief — customers worth attention this morning |
| 2 | aria:churn-watch |
Starter+ | Customers crossing churn-risk thresholds with grounded citations |
| 3 | aria:agency-rollup |
Agency / Enterprise | Cross-brand rollup of KPIs and outliers |
| 4 | aria:dsar-assist |
Starter+ | Drafts a DSAR export bundle for legal review |
| 5 | aria:bench-review |
Starter+ | Reviews recent agent decisions and ranks by review priority |
| 6 | aria:compliance-snapshot |
Growth+ | Current compliance posture with honest hedges |
Resources (6)
aria://customers/{id}aria://conversations/{id}aria://flows/{id}aria://agents/{id}aria://kb/{source_id}aria://analytics/dashboard
Metadata
Release files for aria-cx-mcp 1.9.5.post3
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| aria_cx_mcp-1.9.5.post3.tar.gz | 110.8 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| aria_cx_mcp-1.9.5.post3-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 224.9 kB
Release files / aria_cx_mcp-1.9.5.post3.tar.gz
| Download URL | aria_cx_mcp-1.9.5.post3.tar.gz |
|---|---|
| Size | 110.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
26870819fcc4679275f3f071a353195a95adc8a5f0340b1563d7b8bbbbc1805f
|
|
BLAKE2b-256 checksum How to use checksums |
19fd3f9344c7df361eaf136ac264e22e590fba052c296cd5e45da013e04d4f3c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.3
|
Release files / aria_cx_mcp-1.9.5.post3-py3-none-any.whl
| Download URL | aria_cx_mcp-1.9.5.post3-py3-none-any.whl |
|---|---|
| Size | 114.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
e09e4bb8671dd77c08c0f993bee8e2f8ef4e80250d6d97cc89b909b62785eb78
|
|
BLAKE2b-256 checksum How to use checksums |
0ef343751825c4c2ce2813a67a4dd7009a5659cfb83163eb52962d5e64a8be82
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.3
|