# Reported issues for aviation-weather-mcp-server Pod holds 12 of 12 GitHub reports that passed its relevance review. This can include external user reports, maintainer-confirmed bugs, and concrete feature gaps. Treat them as evidence to inspect, not a count of distinct defects. Back to [aviation-weather-mcp-server](/mcp/aviation-weather-mcp-server). ## Most discussed ### bug(aviation_get_pireps): icing intensity carries concatenated type ### Server version 0.4.1 ### mcp-ts-core version 0.12.3 ### Runtime Bun ### Runtime version Bun 1.3.14 ### Transport stdio ### Description `aviation_get_pireps` passes AWC's `icgInt1` / `icgInt2` through to `icing[].intensity` verbatim. For a large share of reports that value is not a clean intensity code but an intensity concatenated with the literal suffix `clr` — `NEGclr` rather than `NEG` — and it reaches both response surfaces unchanged, rendering as `- NEGclr`. Live sampling… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/26) · 2026-08-13 · closed · 5 comments ### feat(aviation_get_pireps): cap the response size independently of the search ### Use case `aviation_get_pireps` has no per-call size control. The only way to spend fewer tokens is to ask a different question — a smaller box, a shorter window, a narrower altitude band — so a caller who genuinely wants a wide area has no way to sample it cheaply. A CONUS bounding box at `hours=12` draws the upstream maximum of 400 reports every time it is called, and the response runs on the order of 230,000 bytes — roughly 58,000 tokens — with a few percent of drift call to call as the… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/34) · 2026-09-09 · closed · 3 comments ### bug(aviation_find_stations): a stray space in one station_id fails the whole call ### Server version 0.4.2 ### mcp-ts-core version 0.12.8 ### Runtime Bun ### Runtime version Bun 1.4.0 ### Transport http ### Description `aviation_find_stations` accepts each entry of `station_ids` as an unconstrained string and forwards it to the upstream registry untrimmed. An entry carrying leading or trailing whitespace makes the upstream reject the request outright, so a batch of otherwise-valid identifiers returns nothing at all rather than the stations it could have resolved.… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/35) · 2026-09-10 · closed · 2 comments ### feat(aviation_get_advisories): disclose a zero-result draw the way the sibling tools do ### Use case `aviation_get_advisories` is the only tool in this server with no `enrichment` block. A zero-result draw renders as `**0 active advisory(ies)**` in `content[]` and an empty `advisories` array in `structuredContent`, with nothing machine-readable distinguishing "no advisories are active anywhere" from "advisories are active but the filter excluded them all." `aviation_find_stations` and `aviation_get_pireps` both declare a `notice` field and populate it on their zero and truncated… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/33) · 2026-08-25 · closed · 2 comments ### bug(aviation_get_advisories): hazard filter substring-matches an undocumented upstream vocabulary ### Server version 0.3.0 ### mcp-ts-core version 0.11.5 ### Runtime Bun ### Runtime version Bun 1.3.14 ### Transport stdio ### Description `aviation_get_advisories` filters by hazard client-side, with a substring test that assumes the caller's vocabulary and AWC's record vocabulary are the same words: ```ts const hazardUpper = params.hazard.toUpperCase(); advisories = advisories.filter((a) => a.hazard.toUpperCase().includes(hazardUpper)); ``` The direction of that test is what… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/30) · 2026-08-13 · closed · 2 comments ### feat(advisories): serve G-AIRMET and Alaska AIRMET hazard advisories ### Use case The server covers SIGMETs and has no path to AIRMET-family hazards at all. `aviation_get_advisories` reads AWC's `/api/data/airsigmet`, which serves domestic SIGMETs only — its `airSigmetType` field is pinned to `enum: [SIGMET]` in the [AWC OpenAPI schema](https://aviationweather.gov/data/schema/openapi.yaml). A companion bug fixes the dishonest half of that gap by rejecting AIRMET requests with a typed error instead of answering them with convective SIGMETs. This issue is the… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/29) · 2026-08-13 · open · 2 comments ### bug(format): empty cloud array reported as a clear sky ### Server version 0.1.9 ### mcp-ts-core version 0.11.5 ### Runtime Bun ### Runtime version Bun 1.3.14 ### Transport stdio ### Description `aviation_get_metar` and `aviation_get_taf` both render an empty `clouds` array as `**Clouds:** Clear`. That array conflates two different upstream states, and only one of them is a clear sky. AWC does not encode a clear-sky group as a cloud layer — a METAR reporting `CLR` comes back with `clouds: []` and no `CLR` entry. But an observation that… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/27) · 2026-08-13 · closed · 2 comments ### feat(aviation_find_stations): cap the response size independently of the search ### Use case Related: #34 — the same gap on `aviation_get_pireps`, with the same mechanism and the same resolution shape. `aviation_find_stations` has no per-call size control in its area-search modes. A state search returns every station in the state, and a bounding-box search returns everything in the box up to the upstream cap. Neither result is wrong and both disclose truncation correctly, but a caller has no way to ask for less of one. A single-state search (California, 270 stations)… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/36) · 2026-09-10 · closed · 1 comment ## Most recent ### feat(aviation_get_pireps): push altitude and intensity filters upstream of the result cap ### Use case `aviation_get_pireps` filters on the wrong side of the upstream result cap. `altitude_min_ft` and `altitude_max_ft` are applied in the handler, after AWC has already truncated the page at its 400-entry maximum, so they can only subtract from whatever survived truncation. The `pirep` endpoint can filter on altitude itself, before the cap, and filtering before it changes which reports survive. The gap is large enough to change an answer. A CONUS bounding box at `age=12` returns… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/32) · 2026-08-13 · closed · 1 comment ### bug(aviation_find_stations): disclose station IDs that resolved to nothing ### Server version 0.3.0 ### mcp-ts-core version 0.11.5 ### Runtime Bun ### Runtime version Bun 1.3.14 ### Transport stdio ### Description `aviation_find_stations` takes up to 20 station IDs and returns only the ones that resolved. An ID that resolves to nothing is dropped without a trace: the requested list is never echoed, no field names the omission, and the rendered text contains only the stations that came back. Both response surfaces present a partial result as a complete one.… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/31) · 2026-08-13 · closed · 1 comment ### bug(aviation_get_pireps): malformed flight-level group reports altitude 0 ### Server version 0.4.1 ### mcp-ts-core version 0.12.3 ### Runtime Bun ### Runtime version Bun 1.3.14 ### Transport stdio ### Description `aviation_get_pireps` returns `altitude_ft: 0` for a PIREP whose raw flight-level group is present but not a number and not one of the three tokens normalization recognizes as unknown. Zero is a valid-looking altitude, so the report reads as a surface observation rather than one with no usable altitude. Normalization treats `/FLUNKN/`,… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/25) · 2026-08-13 · closed · 1 comment ### feat(aviation_preflight_brief): collect route timing and altitude ### Use case The preflight briefing prompt asks for a flight-window and route-level weather assessment without collecting the time, planned altitude, or route geometry needed to perform one, and frames the result as a Go/No-Go recommendation despite having no pilot, aircraft, or operational-minima context. A practical preflight workflow needs to connect station weather, PIREPs, and advisories to the planned flight rather than to departure and destination identifiers alone. Today the prompt can… [Read the thread](https://github.com/cyanheads/aviation-weather-mcp-server/issues/22) · 2026-08-11 · open · 1 comment The remaining reports are on [the project's issue tracker](https://github.com/cyanheads/aviation-weather-mcp-server/issues).