cli-anything-qgis
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data sources such as QGIS projects (.qgs/.qgz), GeoPackage files, and WKT geometries. This creates a potential surface for indirect prompt injection if the agent is directed to process malicious files provided by untrusted sources.
- Ingestion points:
project open,feature add --wkt,process run --param INPUT=...in SKILL.md. - Capability inventory: The skill possesses file-write capabilities (saving projects, exporting PDFs/images) and executes subprocesses via the
qgis_processruntime. - Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are present in the agent guidance to handle untrusted data.
- Sanitization: There is no explicit evidence of data sanitization or validation instructions for the external content processed by the QGIS runtime.
Audit Metadata