The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local commands for performance benchmarking and data analysis.
Runs cargo run --release -p memory-benchmark to execute the local Rust benchmark crate.
Runs python3 perf/memory/analyze-dhat.py to process local JSON heap profiles generated by the benchmark.
[DATA_EXPOSURE]: The skill deals with memory profiling data (heap allocations, RSS stats) through the dhat and memory-stats libraries. All analysis and output are stored and processed locally within the dhat-heap.json file. No evidence of data exfiltration or access to sensitive credentials (API keys, SSH keys, etc.) was found.
[EXTERNAL_DOWNLOADS]: No external network operations or remote code downloads were identified. All scripts and binaries referenced are part of the local repository structure (perf/memory/).

memory-benchmark