Imagery Provider Adaptation

What it does

When your agent uses it

• : new imagery provider
• Add source type
• Satellite provider
• Imagery ingestion

Example prompts

• “new imagery provider”
• “add source type”
• “satellite provider”
• “/imagery-provider-adaptation”

Requirements

• Python 3

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

1. 1Source Type Definition — Model configuration in hastegeo.core.models
2. 2Download Handler — Provider-specific authentication and download logic
3. 3Band Mapping — Map provider bands to HASTE's expected band order
4. 4Metadata Parser — Extract spatial metadata from provider-specific formats
5. 5Preprocessing Rules — Resolution normalization, radiometric correction
6. 6Tests — Integration tests with sample provider data

What it can do on your machine

Read from SKILL.md and the folder at commit 079bd89. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md (its code samples are python).

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Imagery Provider Adaptation loads about 1.3k tokens when it runs. Until then it costs about 88 tokens; SKILL.md has 367 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~88

When it runs · the whole SKILL.md, loaded when a task matches

~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from microsoft/haste at commit 079bd89, republished under its MIT licence (© microsoft). 367 words, ~1,280 tokens.

name

imagery-provider-adaptation

description

Imagery provider adaptation skill for HASTE. Encapsulates provider-specific logic for satellite imagery sources (Planet, Vantor, Airbus, etc.). Use when: 'new imagery provider', 'add source type', 'satellite provider', 'Planet', 'Vantor', 'Airbus', 'Pleiades', 'WorldView', 'SkySat', 'imagery ingestion', 'provider adapter'.

source

HASTE imagery processing pipeline, satellite provider documentation

domain

geospatial

level

advanced

agents

gis, backend-dev

created_date

2026-04-27

status

draft

# New source type with provider-specific configuration
class NewProviderConfig(BaseModel):
    provider_name: str
    api_url: str
    band_order: list[str]  # e.g., ["B", "G", "R", "NIR"]
    default_crs: str  # e.g., "EPSG:4326"
    tile_size: int  # e.g., 256

# Handle provider-specific authentication and URL patterns
# Use requests with proper auth (API key, OAuth, etc.)
# Stream large files to avoid memory issues
# Validate downloaded file integrity

# Map provider bands to HASTE standard order
# HASTE expects: [Blue, Green, Red, NIR] for 4-band
# Handle extra bands (e.g., coastal, red-edge, SWIR)
# Handle missing bands (e.g., panchromatic only)

# 1. Validate CRS — reproject if needed
# 2. Normalize resolution — resample to target GSD
# 3. Apply radiometric correction if needed
# 4. Generate COG with internal tiling and overviews
# 5. Validate output with rasterio

# Test with real sample data (small AOI, public data preferred)
# Verify CRS preservation
# Verify band order mapping
# Verify COG compliance
# Verify metadata extraction

Format: Cloud Optimized GeoTIFF
Tiling: 256x256 or 512x512 internal tiles
Overviews: Nearest power of 2, down to 256px
Compression: LZW or DEFLATE
CRS: Preserve source CRS (typically UTM or EPSG:4326)
Nodata: 0 for uint8/uint16, NaN for float
Bands: Blue, Green, Red, NIR (minimum)

© microsoft, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .github/skills/imagery-provider-adaptation of microsoft/haste.

Open the folder on GitHubat commit 079bd89