heap-snapshot-analysis
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: The skill includes instructions to 'IGNORE Prior Investigations' and 'Do NOT read, consult, or be influenced by' specific repository and memory folders. While designed to prevent analysis bias, these directives match common patterns used to attempt to override agent context or memory.
- [INDIRECT_PROMPT_INJECTION]: Ingestion points: V8 heap snapshot files containing arbitrary strings from application memory. Boundary markers: The skill does not specify delimiters or warnings for processed heap content. Capability inventory: Local file writing and shell execution via Node.js. Sanitization: No explicit sanitization of memory-derived strings is provided.
- [COMMAND_EXECUTION]: The profiling workflow involves the creation and execution of custom Node.js scripts in a scratchpad directory, which are run through the shell to perform memory analysis.
- [DATA_EXFILTRATION]: Heap snapshots often contain sensitive information such as credentials, PII, and session data. The skill's analysis and logging processes could potentially expose this data in console output or resulting findings reports if not handled with care by the user.
Audit Metadata