nostr-social-graph

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface due to its handling of untrusted external data from the Nostr network.
  • Ingestion points: Untrusted data enters the agent context through event fetching operations such as fetchKind3 and fetchRelayLists described in SKILL.md and references/outbox-model.md.
  • Boundary markers: There are no instructions or delimiters provided to prevent the agent from interpreting embedded commands within fetched strings (e.g., petnames, mute words, or list content).
  • Capability inventory: The skill enables the agent to discover relays and build complex social graphs, which could be manipulated by malicious event content.
  • Sanitization: The implementation patterns do not include sanitization or validation of strings retrieved from the Nostr network before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 11:10 PM