The Agent Skills Directory

[COMMAND_EXECUTION]: The PluginManager class in scripts/src/echecker/plugins/manager.py features a dynamic plugin discovery mechanism. Specifically, the _discover_plugins_in_path method searches for files matching the *_plugin.py pattern and executes them using importlib.util.spec_from_file_location and spec.loader.exec_module. This allows for the execution of arbitrary Python code found within the search paths.\n- [PROMPT_INJECTION]: The skill architecture is vulnerable to indirect prompt injection through the processing of untrusted external data.\n
Ingestion points: Data enters the agent's context through scripts/src/echecker/excel/provider.py (reads Excel cell content) and scripts/src/echecker/rules/v3_parser.py (parses YAML rule files).\n
Boundary markers: Absent. There are no explicit instructions to the agent to disregard instructions potentially embedded within cell data or rule metadata.\n
Capability inventory: The skill can read from and write to the filesystem, including generating annotated Excel files and HTML reports via scripts/src/echecker/excel/annotator.py and scripts/src/echecker/reports/html_reporter.py.\n
Sanitization: Risk is mitigated by the implementation of a custom expression evaluator in scripts/src/echecker/expression/evaluator.py, which performs its own parsing and avoids using dangerous built-in functions like eval() or exec().

e-checker