accint-solve

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates through localized MCP tool calls (acc_retrieve, acc_act) for memory management and does not initiate unauthorized network connections or execute arbitrary shell commands.
  • [PROMPT_INJECTION]: The skill interacts with retrieved memory, which technically creates a surface for indirect prompt injection. However, it includes safety instructions to treat retrieved memories as evidence to be verified against the current repository state rather than as authoritative instructions.
  • [DATA_EXFILTRATION]: The skill instructions include privacy best practices, explicitly directing the agent to avoid recording secrets, credentials, or unnecessary sensitive data in the persistent memory notes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 01:16 PM
Security Audit — agent-trust-hub — accint-solve