analyzing-shellcode

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data, such as raw shellcode blobs, base64 buffers, and PCAP objects. This creates an attack surface for indirect prompt injection if an agent automates these workflows without proper boundary enforcement.
  • Ingestion points: Processes external files including sc.txt, sc.bin, and payload.b64.
  • Boundary markers: The skill does not define specific delimiters to isolate untrusted data from the agent's logic.
  • Capability inventory: The skill utilizes powerful execution and analysis tools such as python3, scdbg, and speakeasy.
  • Sanitization: The skill provides explicit safety guidance, recommending the use of isolated virtual machines (VMs) and hardware emulation to mitigate execution risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 04:46 PM
Security Audit — agent-trust-hub — analyzing-shellcode