openclaw-test-heap-leaks
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill guides the agent to use standard tools like
pnpmandnodefor local performance testing and heap analysis. These actions are appropriate for the skill's stated purpose of memory leak investigation. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external heap snapshot files. Ingestion points: .heapsnapshot files processed by
heapsnapshot-delta.mjs. Boundary markers: Agent instructions specify reporting object families and deltas. Capability inventory: Shell command execution for profiling. Sanitization: The analysis script uses a custom JSON scanner to extract only relevant numeric and type-name fields, effectively preventing the execution of arbitrary instructions embedded in the snapshots. - [SAFE]: The skill correctly identifies the privacy risks of heap snapshots and includes clear instructions to keep these artifacts within a local, controlled scope.
Audit Metadata