The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to install the feature-client Python package directly from the vendor's internal GitLab repository using uv pip install with a pinned commit hash.
[COMMAND_EXECUTION]: The skill utilizes shell commands to verify package installations, manage Python environments, and execute the library's test suite.
[INDIRECT_PROMPT_INJECTION]: The skill processes data fetched from an external feature database (e.g., via get_features) and possesses the capability to write back to the database (e.g., via save_features). This creates a surface where untrusted data from the database could influence agent behavior.
Ingestion points: Data is ingested via get_features, get_satellites, get_feature_types, get_labellers, and get_tags in references/querying.md.
Boundary markers: The skill does not define specific delimiters for separating data content from instructions, though it implements structural boundaries through GeoDataFrames.
Capability inventory: The skill has write access via save_features, create_*, and update_usage_right methods as documented in references/saving-features.md and references/reference-data-and-errors.md.
Sanitization: No explicit sanitization or filtering of database content is described in the provided instructions.

feature-db-client