trading-ledger
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data retrieved from a Notion database during trade reviews, creating an attack surface where instructions embedded in fields like 'Thesis' or 'Plan' could influence agent behavior. 1. Ingestion points: Notion database queries in the 'Batch reconcile and review' step (SKILL.md). 2. Boundary markers: None mentioned for retrieved database content. 3. Capability inventory: Notion tool calls (search, query, create, update). 4. Sanitization: None defined for data retrieved from Notion.
- [EXTERNAL_DOWNLOADS]: The skill references repositories and issue trackers on GitHub, a well-known service. Evidence: Links to the author's repository and the official Notion connector repository are found in the metadata and pitfall sections of SKILL.md.
Audit Metadata