scaffold-rules

Warn

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute system commands, specifically for running the bunx package runner.
  • [REMOTE_CODE_EXECUTION]: The skill invokes bunx @plaited/development-skills, which downloads and executes a remote package from the NPM registry. This package is not from a source listed in the trusted vendors list or the author's known namespace.
  • [EXTERNAL_DOWNLOADS]: Uses bunx to implicitly fetch external code and dependencies from the NPM registry during runtime.
  • [PROMPT_INJECTION]: The skill modifies instruction files used by other AI agents (AGENTS.md and CLAUDE.md), creating a surface for indirect prompt injection. The content generated by the external tool is written into these files, which can influence the behavior of other agents in the environment.
  • Ingestion points: Output from the @plaited/development-skills CLI tool.
  • Boundary markers: Employs <!-- PLAITED-RULES-START --> and <!-- PLAITED-RULES-END --> tags.
  • Capability inventory: The skill has the ability to write to the local filesystem and modify persistent configuration files.
  • Sanitization: No explicit sanitization or validation is performed on the rule content generated by the external package before it is applied to the files.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 28, 2026, 09:25 AM