The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a local shell script scripts/log.sh to manage the decision log. The script performs directory creation using mkdir -p and appends data to a TSV file. It includes a clean function that sanitizes inputs by replacing tabs and newlines with spaces and prepending a single quote to values starting with formula triggers (=, +, -, @), which is a best practice for preventing CSV/spreadsheet injection attacks.
[SAFE]: All operations are local to the workspace. The skill does not perform network operations, access sensitive credentials, or download external code. The instructions promote safe behavior by advising the agent to treat task history as untrusted and to verify evidence against the repository state.

show-me-your-work