geo-intelligence
Warn
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Extensive use of the
subprocessmodule across multiple scripts to orchestrate analysis and interact with system tools. accel.pyrunsnvidia-smi,powershell, anduvto manage GPU acceleration environments.net.py,osm.py,pano.py, andgsv.pyusecurlto fetch data from various external geospatial APIs.recon.pyandphoto_session.pyuseuv runto execute other scripts in the suite.meta.pyexecutesexiftoolto extract deep metadata from images.- [DYNAMIC_EXECUTION]: Several scripts employ dynamic execution patterns for performance or functionality.
textgeo.pyandgeodata.pyusepickle.load()to deserialize local cache files (cities500.v3.pkl), which is unsafe if the cache files are tampered with.accel.pyperforms dynamic environment setup by runninguv pip installto configure a GPU-accelerated environment at runtime.- Static analysis identifies the use of
eval()insat_scan.pyandmatch.pyfor coordinate or geometric processing. - [EXTERNAL_DOWNLOADS]: The skill downloads significant external resources from well-known and research-oriented sources.
- Fetches local ML models (approx. 6.2 GB) from Hugging Face (
StreetCLIP,GeoCLIP,OWLv2, etc.). - Downloads the
GeoCaliblibrary directly from a GitHub repository at runtime. - Fetches geospatial data from OpenStreetMap (Overpass), GeoNames, and Socrata open-data portals.
- [INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface where untrusted data from images is ingested into the agent's context.
- Ingestion points:
ocr.pyandintake.pyextract text from user-supplied images. - Boundary markers: The instructions in
SKILL.mdremind the agent to treat metadata and extracted text as hypotheses, but explicit delimiters for OCR output are not strictly enforced. - Capability inventory: The skill has significant capabilities including network access, file system writes, and subprocess execution across all 38 scripts.
- Sanitization: Extracted text is used to construct queries for
poi.pyandtextgeo.py, creating a path for potential injection if the LLM interprets text as instructions.
Audit Metadata