skills/plurigrid/asi/iroh-p2p/Gen Agent Trust Hub

iroh-p2p

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill introduces a surface for indirect prompt injection by demonstrating how to process untrusted data from a peer-to-peer network.
  • Ingestion points: Untrusted data is fetched via Iroh's blob, document, and gossip services, specifically highlighted in the node.blobs.get_bytes and node.gossip.subscribe snippets in SKILL.md.
  • Boundary markers: The provided examples do not include delimiters or instructions to treat external network data as untrusted or to ignore instructions embedded within it.
  • Capability inventory: The skill demonstrates capabilities that could be exploited if malicious instructions are processed, such as writing data to the local file system using std::fs::write in SKILL.md.
  • Sanitization: No content validation or sanitization logic is present in the code examples to handle potentially malicious payloads from the P2P network.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 07:07 AM