The Agent Skills Directory

[SAFE]: The skill's core logic is implemented in scripts/disambiguate.py, which uses only the Python standard library and involves no network requests, model calls, or dynamic code execution. It processes text requirements through deterministic regular expressions and a custom word-counting algorithm.
[COMMAND_EXECUTION]: The test suite files (tests/mutation_proof.py and tests/test_disambiguate.py) use subprocess.run to execute the skill's own scripts and the pytest framework for validation. These operations are scoped to testing the skill's local logic and do not expose the agent or environment to untrusted shell commands.
[EXTERNAL_DOWNLOADS]: The documentation references the official ASD-STE100 website for users to download reference material manually. The skill itself contains no automated download patterns or remote code execution vectors.
[DATA_EXFILTRATION]: Analysis of the source code confirms no network capabilities or credential harvesting patterns. The skill processes local file content and returns structured findings to the user/agent without external transmission.

disambiguate