Dawarich MCP
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Dawarich MCP show my recent location history"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Dawarich MCP
Disclaimer: This project is a work in progress and was vibe-coded without much of a plan. Treat it as experimental until the API surface and behavior have been reviewed more deliberately.
FastMCP server for Dawarich location history.
The server exposes LLM-friendly tools and resources for maps, stats, places, visits, points, and trip workflows. Dawarich currently does not expose /api/v1/trips, so trip mutation tools return a clear unsupported-capability response until those routes exist.
Configuration
Required:
DAWARICH_BASE_URL: Dawarich base URL, for examplehttp://dawarich.dawarich.svc.cluster.local:3000DAWARICH_API_KEY: Dawarich user API key
Optional:
DAWARICH_AUTH_MODE:bearerorquery, defaultbearerDAWARICH_HOST_HEADER: optional HTTPHostheader for deployments that call an internal service URL while Dawarich only allows the public app hostDAWARICH_FORWARDED_PROTO: optionalX-Forwarded-Protoheader, for examplehttpswhen Dawarich forces HTTPS behind a proxyDAWARICH_TIMEOUT_SECONDS: default30DAWARICH_MAX_PAGE_SIZE: default500MCP_TRANSPORT:stdioorhttp, defaultstdiooutside the containerMCP_HOST: default127.0.0.1MCP_PORT: default8000
Related MCP server: QSS GPS MCP
Local Development
uv --cache-dir /tmp/uv-cache-dawarich-mcp run --extra dev pytest
uv --cache-dir /tmp/uv-cache-dawarich-mcp run --extra dev ruff check .Run over stdio:
DAWARICH_BASE_URL=https://timeline.example.test \
DAWARICH_API_KEY=... \
uv run dawarich-mcp --transport stdioRun over HTTP:
DAWARICH_BASE_URL=https://timeline.example.test \
DAWARICH_API_KEY=... \
uv run dawarich-mcp --transport http --host 127.0.0.1 --port 8000HTTP MCP endpoint: http://127.0.0.1:8000/mcp
Health endpoint: http://127.0.0.1:8000/health
Kubernetes Shape
The intended cluster deployment is internal only:
Namespace:
dawarichService:
dawarich-mcpURL:
http://dawarich-mcp.dawarich.svc.cluster.local:8000/mcpNo IngressRoute
Dawarich API URL:
http://dawarich.dawarich.svc.cluster.local:3000
Store the API key in a Kubernetes Secret. Do not commit it.
automated-setups expects the secret at:
namespace: dawarich
name: dawarich-mcp-secret
key: api-keyThe image build and deployment can be run separately:
# Build the configured localhost/dawarich-mcp image into Spark's k8s.io containerd namespace.
direnv exec . env ANSIBLE_LOCAL_TEMP=/tmp/ansible-local TMPDIR=/tmp \
venv/bin/ansible-playbook -i inventory/k8s_cluster/inventory.ini \
playbooks/cluster/applications.yml --tags dawarich-mcp-build
# Apply the internal-only Deployment and ClusterIP Service.
direnv exec . env ANSIBLE_LOCAL_TEMP=/tmp/ansible-local TMPDIR=/tmp \
venv/bin/ansible-playbook -i inventory/k8s_cluster/inventory.ini \
playbooks/cluster/applications.yml --tags dawarich-mcp-deployThe deploy tag refuses to apply the Deployment unless either
dawarich_mcp_api_key is set from Ansible Vault or the dawarich-mcp-secret
already exists in the dawarich namespace.
Safety
Bearer auth is preferred so the Dawarich API key is not placed in URLs.
Delete, merge, bulk update, and recalculation tools default to dry-run.
Raw point reads are available but described as lower-level tools; prefer timeline, stats, and map context first.
Trip tools intentionally do not scrape Dawarich HTML forms. They report the missing
/api/v1/tripsAPI instead.
Available Tools
19 toolsdawarich_add_pointsA
Add raw Dawarich points. Prefer visit/timeline tools unless raw samples are needed.
Inputs: points (required), dry_run. Returns: object
dawarich_bulk_update_visit_statusA
Bulk update Dawarich visit status. Defaults to dry-run.
Inputs: status (required), visit_ids (required), dry_run. Returns: object
dawarich_create_placeC
Create a manual Dawarich place.
Inputs: name (required), latitude (required), longitude (required), note, +2 more. Returns: object
dawarich_create_tripB
Create a Dawarich trip when /api/v1/trips exists; currently reports that API gap.
Inputs: name (required), ended_at (required), started_at (required), dry_run, +1 more. Returns: object
dawarich_create_visitC
Create a Dawarich visit and matching/manual place if needed.
Inputs: name (required), ended_at (required), latitude (required), longitude (required), +3 more. Returns: object
dawarich_delete_placeB
Delete a Dawarich place. Defaults to dry-run.
Inputs: place_id (required), dry_run. Returns: object
dawarich_delete_tripC
Delete a Dawarich trip when /api/v1/trips exists; currently reports that API gap.
Inputs: trip_id (required), dry_run. Returns: object
dawarich_find_pointsA
Find raw Dawarich points. Prefer timeline/map context unless point evidence is needed.
Inputs: page, slim, order, end_at, +4 more. Returns: object
dawarich_find_visitsA
Find Dawarich visits by time and status.
Inputs: page, end_at, status, per_page, +1 more. Returns: object
dawarich_get_map_contextA
Get map bounds and optional capped hexagon/fog context.
Inputs: end_at (required), start_at (required), include_fog, max_features, +1 more. Returns: object
dawarich_get_statsB
Get aggregate Dawarich travel stats, optionally filtered to one year.
Inputs: year, include_monthly. Returns: object
dawarich_get_timelineC
Get day-by-day Dawarich timeline context for a date range.
Inputs: end_at (required), start_at (required), distance_unit. Returns: object
dawarich_merge_visitsC
Merge Dawarich visits. Defaults to dry-run because merge is destructive.
Inputs: visit_ids (required), dry_run. Returns: object
dawarich_recalculateC
Queue supported Dawarich recalculations. Defaults to dry-run.
Inputs: kind (required), dry_run. Returns: object
dawarich_search_placesC
Search saved and nearby Dawarich places.
Inputs: limit, query, latitude, longitude, +1 more. Returns: object
dawarich_select_visit_placeC
Select one of Dawarich's possible places for a visit.
Inputs: place_id (required), visit_id (required), dry_run. Returns: object
dawarich_update_placeB
Update a Dawarich place after reading its current value.
Inputs: place_id (required), name, note, dry_run, +3 more. Returns: object
dawarich_update_tripC
Update a Dawarich trip when /api/v1/trips exists; currently reports that API gap.
Inputs: trip_id (required), name, dry_run, ended_at, +2 more. Returns: object
dawarich_update_visitC
Update a Dawarich visit after reading its current value.
Inputs: visit_id (required), name, status, area_id, +4 more. Returns: object
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 19 tools
Most tools are clearly distinct: places, visits, points, timeline, stats, and recalculate all have separate purposes. However, the three trip tools (create/update/delete) are functionally identical stubs that all report an API gap, making them hard to distinguish by action despite different names.
All tools follow a consistent dawarich_<verb>_<noun> pattern in snake_case, with verbs like search, find, create, update, delete, add, merge, and bulk_update. This is highly predictable and readable.
At 19 tools, the server is slightly over the ideal 3-15 range but still reasonable for a domain covering places, visits, points, trips, stats, timeline, and map context. The three non-functional trip tools could be trimmed, but the count is not excessive.
Core place/visit/point operations are well represented, plus useful extras like timeline, stats, and map context. However, trip tools are non-functional stubs that only report an API gap, creating dead ends for trip workflows, and there is no delete_visit tool. These are notable gaps.
Maintenance
Related MCP Connectors
Routing, geocoding, places, matrix and map tools for location-aware AI workflows.
Connect geographic IDs, places, shapes, and published data through DaedalMap tools.
Geocode, reverse-geocode, autocomplete, route and search places.
Geocoding, weather forecasts, and timezone lookups
Related MCP Servers
- AlicenseBqualityBmaintenanceEnables tracking and analyzing development time through the Wakapi API. Provides tools to retrieve coding statistics, project details, leaderboards, and recent activity logs for productivity insights.81Apache 2.0
- FlicenseNot gradedqualityDmaintenanceEnables adding GPS goals and logging time directly from Claude Desktop, with support for bulk backfill and session management.-
- FlicenseNot gradedqualityDmaintenanceExposes read-only location history and related data from Dawarich via MCP tools, designed to run behind an OAuth-terminating proxy.-
- FlicenseAqualityDmaintenanceMinimal read-only MCP server that lets you query a Dawarich instance for stats, timeline, visits, points, track details, exports, reverse geocoding, and journey summaries.91-