Nasa Firms Fire Fetch
Retrieve bounded NASA FIRMS satellite active-fire and thermal-anomaly point detections through th…
by tiangong-ai · MIT
The skill
Retrieve bounded NASA FIRMS satellite active-fire and thermal-anomaly point detections through the Tiangong CLI for one reviewed source, known bounding box, and short UTC date window. Use for hotspot screening or candidate fire-location context; do not use for fire perimeters, burned area, confirmed wildfire incidents, cause attribution, emergency alerts, recurring monitoring, or multi-source fusion.
NASA FIRMS Active Fire
Use the CLI-owned nasa-firms.active-fire capability. This Skill supplies
intent routing and result-use boundaries only; the CLI owns source discovery,
input/output schemas, MAP_KEY injection, chunking, limits, validation, and
receipts.
Before running
- Read
references/tiangong-data-requirement.json. - Use the caller- or workspace-resolved stable CLI. The requirement declares compatible capability and operation contract majors; it does not select a package build.
- Run
data describewith that same CLI. Continue only when the capability ID and required contract majors match, and copy the exact current capability/operation versions from that response into the run request. - Ensure
NASA_FIRMS_MAP_KEYis present in the CLI process environment, then run the default static doctor. Never place the key in argv, the request JSON, a Skill-local config file, logs, or output.
tiangong-ai data describe nasa-firms.active-fire --json
tiangong-ai data doctor nasa-firms.active-fire --json
Use the returned Discovery Metadata to confirm current source availability,
coverage, granularity, NRT or Standard Processing semantics, citation guidance,
quota, provides, and doesNotProvide. Do not substitute facts remembered
from an older Skill revision. A blocked static doctor means the required
logical credential is unavailable; stop instead of bypassing the CLI.
Prepare the request
Build a tiangong.data.run-request.v1 envelope and replace the two version
placeholders with the exact versions from the same data describe response. Use one source and one known WGS84
bounding box that does not cross the antimeridian. The inclusive UTC date
window may contain at most 31 dates; the CLI performs provider-compliant
five-day chunking and enforces transaction and record limits.
{
"schemaVersion": "tiangong.data.run-request.v1",
"capabilityId": "nasa-firms.active-fire",
"capabilityVersion": "<describe.manifest.capabilityVersion>",
"operationId": "fetch-area",
"operationVersion": "<describe.manifest.operations[0].operationVersion>",
"input": {
"source": "VIIRS_NOAA20_NRT",
"boundingBox": {
"west": 115.8,
"south": -8.9,
"east": 116.3,
"north": -8.3
},
"startDate": "2026-03-01",
"endDate": "2026-03-07",
"checkAvailability": true
}
}
Use the operation input schema returned by the same data describe response for the current
source enum and limits. Select NRT only when timeliness is material and its
provisional status can be retained; select the matching Standard Processing
source for consistent historical work when available. Set
checkAvailability when the source's historical window is uncertain, accepting
the extra provider request. Do not silently widen the box, switch sources,
geocode place names, cross the antimeridian, or fan out across areas.
Run
tiangong-ai data run nasa-firms.active-fire fetch-area \
--input /absolute/path/to/request.json --json
The command emits a tiangong.data.run-result.v1 envelope. Preserve its
contract, warnings, errors, and receipt with data when handing the
result to another workflow.
Result boundaries
- Treat each record as a satellite thermal-anomaly or active-fire detection, not a fire perimeter, burned-area estimate, incident identity, or proof of a wildfire or ignition cause.
- Preserve source, acquisition time, satellite/instrument, confidence, day/night, version, and sensor measurements with every downstream use.
- Preserve NRT provisional warnings and the exact selected dataset in citations. Do not merge NRT and Standard Processing records without an explicit method.
- Treat
partialas incomplete chunk or row coverage. Report missing chunks and invalid paths together with the usable detections. - Preserve duplicate/inconsistent-header and row-validation issues. Rows with invalid required coordinates or acquisition timestamps are omitted, while a duplicate optional header or malformed optional measurement does not discard an otherwise usable detection.
- Treat
blockedas no usable business result. Surface credential, request, availability, endpoint, quota, or provider errors instead of bypassing limits or switching endpoints. - Report record-limit truncation and transaction estimates; a truncated result is not exhaustive evidence of hotspot absence or presence.
- Confirm operationally important detections with incident, perimeter, imagery, or local-authority sources. This capability supplies no alert, evacuation, severity, containment, smoke, emissions, weather, or hydrology decision.
- Recurring monitoring, multi-area or multi-source fan-out, duplicate-event resolution, statistical inference, and research evidence admission belong to the caller or Auto Research, not this atomic Skill.
- This execution contract does not expose the source script's standalone MAP_KEY-status probe, arbitrary sensor-specific raw columns, or raw JSON/log artifact output. Use the static doctor plus an explicit bounded operation; add any future live quota diagnostic as a separately reviewed CLI contract.
Reference
references/tiangong-data-requirement.json: stable capability requirement; it is not a package lock.
Use this skill
Connect your agent to the Skill Store MCP server and ask for it by ref: tiangong-ai/nasa-firms-fire-fetch-79455926116.
Details
- Publisher
- tiangong-ai
- Category
- Data & Analytics › Scraping & extraction
- License
- MIT
- Updated
- 2026-09-14
References
More from tiangong-ai
- Academic Paper Download Fetch and atomically save structurally and identity-verified academic-paper PDFs from legal open-…
- Ai Tech Fulltext Fetch Fetch and persist article full text for RSS entries already stored in SQLite by ai-tech-rss-fetch.
- Ai Tech Rss Fetch Subscribe to AI and tech RSS feeds and persist normalized metadata into SQLite using mature Pytho…
- Ai Tech Summary Retrieve time-windowed RSS evidence from SQLite and let the agent produce final summaries using R…