context-bankruptcy
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes historical memory data and session logs, which constitutes an attack surface for instructions that may have been embedded in past interactions.
- Ingestion points: Historical memory contents, exported memory files, and pasted context dumps as defined in the 'Required Inputs' section.
- Boundary markers: The skill suggests converting memory into discrete claims within markdown tables but lacks explicit delimiters or instructions to ignore commands embedded in the source text.
- Capability inventory: The skill is restricted to generating text artifacts (belief audits, ledgers, and records) and does not utilize tools for file system modification, shell execution, or network communication.
- Sanitization: There are no instructions for sanitizing or escaping the content of the memory dump before processing.
- [SAFE]: The skill consists entirely of instructional markdown and does not include any executable code, scripts, or remote dependencies. All operations are designed to be performed within the text context of the agent.
Audit Metadata