heap-snapshot-analysis
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- Indirect Prompt Injection Surface: The skill processes external
.heapsnapshotfiles, which may contain untrusted data from an application's heap. This creates a potential surface where strings embedded in the heap could influence the agent's behavior during analysis.\n - Ingestion points: Snapshot files are read via
helpers/parseSnapshot.tsandhelpers/streamSnapshot.mjs.\n - Boundary markers: No explicit markers are defined to delimit untrusted heap strings from the agent's instructions.\n
- Capability inventory: The skill performs file system operations and executes shell commands for script-based analysis.\n
- Sanitization: Heap-derived strings are not specifically sanitized before being analyzed by the agent.\n- Dynamic Script Execution: The skill encourages the agent to generate and execute JavaScript analysis scripts in a local
scratchpad/directory. While necessary for complex heap investigations, the execution of runtime-generated code is a pattern that warrants review to ensure the generated logic remains within expected bounds.
Audit Metadata