skills/testland/qa/codeql-queries/Gen Agent Trust Hub

codeql-queries

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches the official CodeQL CLI binary from GitHub's releases repository (github.com/github/codeql-cli-binaries).
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for installation (unzip, export) and tool operation. For compiled languages, it requires wrapping the project's build process (e.g., ./gradlew build or make all) with the CodeQL database creation command, which involves executing arbitrary scripts found in the target repository.
  • [PROMPT_INJECTION]: An indirect prompt injection surface is identified where malicious instructions could be embedded in the codebase being processed.
  • Ingestion points: Reads and parses source code from the local repository directory via the --source-root parameter during database creation and analysis. (SKILL.md)
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard natural language prompts or instructions embedded within the scanned source code. (SKILL.md)
  • Capability inventory: Includes shell command execution for both the CodeQL CLI and repository-specific build commands (e.g., ./gradlew build). (SKILL.md)
  • Sanitization: The process relies on CodeQL's internal semantic parsing; no explicit sanitization or filtering of code content is performed before the agent processes the repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 10:54 AM
Security Audit — agent-trust-hub — codeql-queries