The Agent Skills Directory

[SAFE]: The skill contains no executable code, remote dependencies, or network operations. It provides clear, human-readable instructions for a debugging workflow.
[COMMAND_EXECUTION]: The workflow involves instructions to clear build caches and run the project to observe log output. These are standard developer operations necessary for the skill's stated purpose of tracing system flow and do not represent a security risk in this context.
[INDIRECT_PROMPT_INJECTION]: While the skill involves reading and interpreting existing code (an external ingestion surface), the instructions explicitly restrict the agent to observation and instrumentation without applying fixes or altering semantics, mitigating the risk of executing malicious instructions found in the code being analyzed.

trace-flow