secure-workflow-guide

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill guides the agent and user in executing various smart contract security analysis tools. This includes the Slither static analyzer and its utility tools (slither-check-upgradeability, slither-check-erc, slither-prop), as well as the Echidna fuzzer and Manticore symbolic execution engine. These commands are legitimate and necessary for the skill's primary function of conducting a multi-step security workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze arbitrary smart contract source code provided by users. This creates a surface for indirect prompt injection where malicious instructions could be embedded within code comments or string literals to manipulate the agent's findings or behavior.
  • Ingestion points: The skill reads the local codebase to explore its structure and perform security scans (as described in SKILL.md).
  • Boundary markers: The instructions do not explicitly mandate the use of boundary markers or instructions for the agent to ignore embedded commands within the source files.
  • Capability inventory: The skill has the capability to run shell commands and generate report artifacts (diagrams, configs) based on the analyzed code.
  • Sanitization: The skill relies on the established logic of the security tools (like Slither) to process the code, which serves as a functional layer of sanitization for security analysis purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 09:17 PM
Security Audit — agent-trust-hub — secure-workflow-guide