fast-io
Pass
Audited by Gen Agent Trust Hub on Apr 30, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it is designed to ingest, summarize, and perform RAG (Retrieval-Augmented Generation) on external documents provided by users.
- Ingestion points: Document and code file ingestion into workspace storage, as described in
REFERENCE.md. - Boundary markers: No specific boundary markers or instructions to ignore embedded prompts are documented.
- Capability inventory: The skill possesses
aichat capabilities, RAG indexing, storage management, and metadata extraction tools (README.md,REFERENCE.md). - Sanitization: No sanitization or validation of the processed document content is mentioned.
- [EXTERNAL_DOWNLOADS]: The documentation provides instructions to install the official Fast.io CLI tool using the NPM registry (
@vividengine/fastio-cli). This is a standard vendor-provided dependency. - [REMOTE_CODE_EXECUTION]: The documentation includes a command for installing the vendor's CLI tool by piping a script from their official GitHub repository (
raw.githubusercontent.com/MediaFire/fastio_cli/main/install.sh) directly into the shell. While this is the vendor's own installation method, it involves the execution of remote code via a piped shell command.
Audit Metadata