duckdb
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPRIVILEGE_ESCALATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download the DuckDB CLI binary from its official releases on GitHub (
github.com/duckdb/duckdb). This is a well-known and recognized service for distributing software releases. - [PRIVILEGE_ESCALATION]: The CLI installation guide includes a command using
sudoto move the downloaded binary into a system-wide executable path (/usr/local/bin/). This is a standard procedure for global CLI tool installation. - [COMMAND_EXECUTION]: The skill includes shell commands for environment setup and tool installation, such as
wget,unzip,chmod +x, andmv. These are necessary steps for the documented installation process. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and query external data files, which presents a surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: The skill reads from various file types including
orders.csv,raw_events.parquet,logs.jsonl,users.csv, and glob patterns likedata/events/*.parquet(documented in SKILL.md). - Boundary markers: There are no explicit delimiters or specific instructions for the agent to ignore natural language instructions that might be embedded within the data files.
- Capability inventory: The skill possesses the capability to execute shell commands (via the
duckdbCLI), as well as run code via Python and Node.js environments. - Sanitization: The examples show direct querying of file paths without explicit validation or sanitization of the input file sources.
Audit Metadata