Skills
Navigate
skills/dslsdzc/rev-skills/re-wasm/Gen Agent Trust Hub

re-wasm

Fail

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill includes an installation command curl https://wasmtime.dev/install.sh -sSf | bash that downloads and executes a script from the official Wasmtime project domain.
  • [INDIRECT_PROMPT_INJECTION]: The skill involves processing untrusted binary modules which are converted to text and interpreted by the agent, creating a vulnerability to indirect instructions.
  • Ingestion points: The agent reads the results of disassembly and decompilation tools such as wasm2wat, wasm-objdump, and wasm-decompile (SKILL.md).
  • Boundary markers: There are no specified boundary markers or instructions to isolate the output of these tools from the agent's reasoning process.
  • Capability inventory: The skill allows for command execution via wasmtime, wasm3, and node, as well as starting a local server with python3 -m http.server (SKILL.md).
  • Sanitization: The instructions do not include steps to sanitize or filter the content of the WebAssembly modules being analyzed.
  • [DYNAMIC_EXECUTION]: The skill instructs the user/agent to run untrusted WebAssembly binaries using the wasmtime and wasm3 runtimes, which involves executing arbitrary code in a sandboxed environment.
Recommendations
  • HIGH: Downloads and executes remote code from: https://wasmtime.dev/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 14, 2026, 09:10 AM