fable-memory
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to create persistent memory records that influence agent behavior across sessions.
- Ingestion points: The
memory_factinput inSKILL.mdis the primary entry point for untrusted data. - Boundary markers: While the skill uses structured markdown templates (
templates/memory-fact.template.md), it lacks explicit technical delimiters to prevent the agent from interpreting instructions embedded within the facts. - Capability inventory: The skill has workspace mutation capabilities (
mutatesWorkspace: true) for creating and updating fact files in.fable/memory/facts/and theMEMORY.mdindex. - Sanitization: The protocol relies on instructions to exclude sensitive data and revalidate facts but does not implement technical sanitization to prevent the storage of adversarial instructions disguised as facts.
Audit Metadata