halo-asset-context

Halo Asset Context

Review an asset in Halo together with the tickets and CMDB/CI relationships around it — entirely read-only. Halo is HaloPSA / HaloITSM (the same product and REST API); assets are called "Devices" in Halo's API vocabulary. Use this to build a full operational picture of a device before acting on it, correlating its history and dependencies.

MCP Server

Server: halo-mcp (NetClaw-authored, mcp-servers/halo-mcp/)
Command: python3 -u mcp-servers/halo-mcp/halo_mcp_server.py (stdio transport)
Auth: OAuth2 client-credentials — HALO_CLIENT_ID / HALO_CLIENT_SECRET against HALO_BASE_URL
Python: 3.10+ · Dependencies: fastmcp, httpx, python-dotenv
Read/write posture: The whole server is read-only except one gated write (halo_create_change_request, in the halo-change-request skill). Every tool in this skill is read-only and makes no changes to Halo.

Available Tools

Tool	Parameters	What It Does
`halo_list_assets`	`customer?`, `assettype_id?`, `search?`	List/search assets ("Devices"), optionally scoped by customer, asset type, or free text
`halo_get_asset`	`asset` (name / inventory-number / id)	Read one asset — detail, fields, and ticket counts
`halo_get_asset_tickets`	`asset`, `open_only?`	List the tickets related to an asset — the core context tool
`halo_get_asset_relationships`	`asset`	An asset's CMDB/CI hierarchy and relationship context
`halo_list_clients`	`search?`	Resolve a Halo Client name to its id (the `customer` param)
`halo_list_sites`	`customer?`, `search?`	Narrow to a site within a client

Key Concepts

Halo == HaloPSA == HaloITSM. Same API for both editions.
Assets are "Devices." halo_list_assets / halo_get_asset accept an asset name, inventory number, or numeric id — you do not need the id up front.
The Halo "Client" is the customer. Scope asset queries to an org with the customer param (a client name or id); the server resolves it to a client_id. In an MSP (HaloPSA) tenant this is essential to avoid crossing client boundaries.
Two context dimensions: tickets (what has gone wrong / changed on this device) via halo_get_asset_tickets, and relationships (what this device depends on or supports) via halo_get_asset_relationships. Use both for a complete picture.
Read-only. Nothing here writes to Halo. If the review concludes a change is needed, hand off to halo-change-request.

Workflow

Review an asset and everything around it:

Locate the asset. If you only have a description or partial name, search first:
```
halo_list_assets(customer="<client>", search="<name/serial/model>")
```
Scope by customer (the Halo Client) in multi-client tenants to disambiguate.
Read the asset detail:
```
halo_get_asset(asset="<name / inventory number / id>")
```
Note its type, key fields, owning client/site, and ticket counts.

Pull its related tickets (the core step):

halo_get_asset_tickets(asset="<asset>")               # all history
halo_get_asset_tickets(asset="<asset>", open_only=true) # active issues only

Look for recurring faults, recent changes, and open work.

Map its CMDB/CI relationships:
```
halo_get_asset_relationships(asset="<asset>")
```
Understand upstream dependencies and downstream impact before anyone touches it.
Report an asset-context summary: identity, owning client/site, open vs historical tickets, notable recurring issues, and CI dependencies — plus a recommendation (e.g. "raise a change via halo-change-request" or "no action").

Integration with Other Skills

Skill	Integration
gait-session-tracking	Mandatory. Record every asset/ticket/relationship query in the GAIT audit trail
halo-change-request	Hand off here when the review concludes a change is warranted — pass the affected asset into the `asset` param
halo-ticket-context	Drill into any individual ticket surfaced by `halo_get_asset_tickets` for its full action history
nautobot-sot / netbox-reconcile	Cross-reference Halo's asset record against the network source of truth; ticket any drift
memory	Recall prior findings about the same device across sessions

Environment Variables

Variable	Required	Description
`HALO_BASE_URL`	Yes	Halo host, e.g. `https://<tenant>.halopsa.com`
`HALO_CLIENT_ID`	Yes	OAuth2 client-credentials application id
`HALO_CLIENT_SECRET`	Yes	OAuth2 client secret
`HALO_TENANT`	Optional	Tenant identifier
`HALO_SCOPE`	Optional	OAuth2 scope (default `all`)
`HALO_AUTH_URL`	Optional	Override the auth-server URL for self-hosted layouts
`HALO_VERIFY_SSL`	Optional	`true`/`false` TLS verification (default `true`)
`HALO_TIMEOUT`	Optional	Per-request timeout in seconds (default 30)
`HALO_PAGE_SIZE` / `HALO_MAX_PAGES`	Optional	Pagination tuning (defaults 50 / 20)
`HALO_RATE_LIMIT`	Optional	Requests/minute cap (0 = disabled)

Important Rules

Read-only skill. None of these tools change Halo. The only write in the entire halo-mcp server is halo_create_change_request (halo-change-request skill).
Scope by the Halo Client. Pass the customer org via the customer param to keep queries inside client boundaries — critical in MSP (HaloPSA) tenants.
Assets resolve by name, inventory number, or id — no need to look up the numeric id first.
Halo == HaloPSA == HaloITSM — identical behavior for both editions.
GAIT logging is mandatory for all queries.
Escalate, don't mutate. If the context implies a change, route it through halo-change-request (preview -> confirm -> submit); never try to write from this skill.

Failure Behavior

If a tool call fails with an authentication or connection error, check that HALO_BASE_URL, HALO_CLIENT_ID, HALO_CLIENT_SECRET are set and valid before assuming a data or device problem.
On a tool error (timeout, unreachable host, malformed response), report the failure and its error message directly to the user rather than fabricating or guessing at results.
For a confirmed read-only call, check connectivity and retry once if appropriate. For any call that changes state or sends a message, a timeout does not prove the action failed: inspect current state or delivery status before retrying, preserve the required approval/change gates, and do not repeat an action whose outcome is unknown.