Peaks MCP server

• Registry namefyi.peaks/mountainsverified in the MCP registry

The registry name sits under peaks.fyi. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

• Stateanswerschecked 2026-10-30
• Tools15
• Server calls itselfpeaks.fyi
• Server version3.1.0
• Protocol version2025-06-18

On 2026-10-30 the server answered initialize and tools/list without a sign-in. First answer in 79 ms.

• Its tools do every use case we checked: filter European peaks by elevation, prominence and difficulty; compare peaks by height and climbing difficulty grade.
• Checked on 6 of the 7 days it takes to judge reliability.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

• INITE MCP scoreafter a weeka score needs 7 daily checks: 6 so far
• What an agent can do with it2 of 2 use cases
• Answeredmeasuring6 of the 7 daily checks it takes to print a share
• Median first answer300 ms
• Space in an agent's context≈ 4,118 tokensthe definitions of all its tools (15), which an agent carries on every turn
• Tool definitionsA · 90 of 100
• Typed results15 of 15 toolsdeclared outputSchema: results come typed, read once

• Parameters without a description: in 9 of 15

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

• get_route_briefs

  Compare up to six known routes in one compact evidence packet: grade, timing conflicts, selected equipment, linked huts, access, notices and optional dated daylight. Preserves per-route errors and source provenance. Partial results must…

• get_capabilities

  Inspect actual catalogue coverage, output JSON schemas, data version, units and unsupported operations before planning. Unknown data is not evidence of absence.

• get_notices

  Read dated notices with publication validity, explicit scope uncertainty and recurrence hints. Supply slug and the trip date; omitted date means today (UTC), echoed in requestedDate. Expired notices may describe recurring regulations…

• resolve_peak

  Resolve a canonical peak ID, external Wikidata/OSM ID, current slug or exact name to its current slug and citable URL. Example input: {"id":"Ben Nevis"}. Use canonical IDs to recover from renamed slugs.

• get_changes

  Compare a watched set of peaks with a previous snapshot. Returns content changes, removals and an updated snapshot. No changes means no new source data, not proof that conditions are unchanged. Date-dependent notice status must still be…

• get_daylight

  Compute sunrise, sunset and civil twilight for the exact calendar date and IANA zone. UTC instants handle daylight saving and polar days. Terrain and weather can reduce usable daylight.

• search

  Search peaks.fyi by mountain name, alternative name, country, range or climbing request. Returns citable peak IDs and URLs. Use fetch to read a result; use search_peaks for explicit filters and pagination. An empty result means no match…

• fetch

  Read a peak search result by catalogue slug as a citable text document. Example input: {"id":"ben-nevis"}. This retrieves catalogue documents, not external website URLs; use your own browsing tool for publisher pages. Includes routes…

• get_peak

  Get compact peak facts and route IDs. Example input: {"slug":"ben-nevis"}. Optionally request access, huts, seasons, hazards or provenance sections. For fact questions search may already suffice. For current weather use get_weather; for…

• get_route

  Get one route by peak slug and routeId from get_peak: grade, equipment, key passages, stages, trailhead, timing, confidence and geometry QA. Failed, truncated or outline geometry must not be used for navigation. Times are estimates. No…

• get_route_point

  Look up a named waypoint or difficult passage on a route. Returns source notes, known coordinates, nearby stage names, key passages and relevant equipment. Missing coordinates or conditions remain unknown; this is not live navigation…

• get_huts

  Read huts, documented seasonal operation, source years and booking/contact URLs for a peak. Beds means capacity, never vacancies. Live availability, reservations, rental inventory and current opening status are unknown. Do not infer next…

• search_routes

  Find routes satisfying all supplied constraints on the SAME route, including difficulty, glacier travel, style and computed round-trip hours. Unknown glacier status is excluded by glacier filters. Searches all candidate peaks; use…

• get_weather

  Get published weather for a peak and optional summit/hut/trailhead target. Set targetId and start/end for a bounded planning response; unfiltered forecasts can be large. Point samples are separate from accumulated rain/snow intervals and…

• search_peaks

  Find and compare mountains with explicit country, elevation, season, difficulty and glacier filters. All filters are hard constraints; never silently relaxed. Returns compact peak summaries, IDs, total and nextOffset. Read get_peak and…