The Agent Skills Directory

[SAFE]: The skill is designed for local project troubleshooting. It uses Python scripts to verify the presence of required project files (e.g., pubspec.yaml, main.dart) and build artifacts.
[COMMAND_EXECUTION]: The skill executes provided local Python scripts (scripts/stac_doctor.py and scripts/check_build_outputs.py) to automate health checks. These scripts are restricted to filesystem read operations within the user-provided project root and do not perform network requests or write operations.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it reads the content of project files (Dart source code, YAML configurations) to perform diagnostics.
Ingestion points: scripts/stac_doctor.py reads .dart files, pubspec.yaml, and .gitignore from the provided project root.
Boundary markers: None identified in the script logic; however, the data is used for status reporting rather than being directly interpolated into high-privilege instructions.
Capability inventory: The skill is limited to reading local files and reporting status via stdout. It lacks network access, file write capabilities, or subprocess spawning of untrusted inputs.
Sanitization: The scripts use specific regex patterns (e.g., searching for @StacScreen) and existence checks, which limits the impact of arbitrary content within the analyzed files.

stac-troubleshooter