memory-integrity-checker
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from an external state file and has the capability to modify that data, creating a potential surface for indirect prompt injection.
- Ingestion points: The script
integrity.pyreads from~/.openclaw/skill-state/memory-dag-compactor/state.yamlvia theload_dag_state()function. - Boundary markers: No explicit delimiters or boundary markers are used to isolate the data being processed from the instructions of the skill.
- Capability inventory: The skill possesses file-write capabilities through the
save_dag_state()function, which can overwrite the compacting skill's state file with modified data. - Sanitization: The skill correctly utilizes
yaml.safe_load()for deserialization, preventing arbitrary code execution during parsing. However, it does not sanitize the logical content of the memory nodes beyond checking structural properties like token counts and depth consistency.
Audit Metadata