static-site-experience-release

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute standard Git operations (commit, merge, push) and repository-specific build and check commands, which are necessary for the deployment workflow.\n- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it processes content from external, potentially untrusted sources to influence site updates.\n
  • Ingestion points: Workflow 1 and 2 involve reading data from the existing static site, verified source files, and user-provided authoritative documents like official PDF resumes.\n
  • Boundary markers: Absent. The instructions do not provide specific delimiters or ignore-instructions to isolate data processed from these files.\n
  • Capability inventory: The agent can modify local source code, execute arbitrary shell commands via the project's build and test scripts, and push changes to the production repository (Workflow 5 and 6).\n
  • Sanitization: Absent. There are no instructions to sanitize or validate the content of ingested files before they are used to update the site structure or metadata.\n- [SAFE]: The skill includes strong defensive instructions, such as explicitly forbidding the introduction of databases or secret-bearing client implementations. It also mandates that all Material changes and deployments require explicit user approval and manual verification steps, which provides a significant safeguard against automated misuse.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 08:08 AM
Security Audit — agent-trust-hub — static-site-experience-release