loose-ends

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill consists exclusively of instructions for the agent to review and summarize its own conversation history. It does not contain any executable code, script references, or network operations.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process the full history of the current conversation, which is an untrusted data source.
  • Ingestion points: The agent is instructed to re-read the entire conversation from oldest to newest (SKILL.md).
  • Boundary markers: The instructions do not specify any delimiters or safety markers to isolate previous user input from the skill's logic.
  • Capability inventory: The skill explicitly restricts the agent to a 'recall-only' mode, forbidding tool use, file writes, or task creation, which effectively negates the risk of an injection causing harmful actions.
  • Sanitization: No sanitization or filtering of the conversation content is performed before the agent lists the unresolved items.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 11:19 PM
Security Audit — agent-trust-hub — loose-ends