The Agent Skills Directory

[COMMAND_EXECUTION]: The skill initiates a local background process via a provided Python script (log-server.py) to collect debugging logs during runtime.- [COMMAND_EXECUTION]: Utilizes standard system utilities like lsof and kill for process management and lifecycle control of the log collection server.- [SAFE]: The log collection service is strictly bound to the local loopback interface (127.0.0.1), ensuring that no debugging data is exposed over the network.- [SAFE]: The skill instructions include explicit prohibitions against logging sensitive information such as passwords, tokens, API keys, or personally identifiable information (PII).- [SAFE]: A mandatory cleanup protocol is defined to ensure all instrumentation code is removed and background processes are terminated once the debugging session concludes.- [SAFE]: Operates on a designated local directory (~/.codeflicker) for data storage, maintaining isolation from sensitive system files.

hypothesis-debugging