ai-readiness-checklist

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the Metabase instance which could contain instructions intended to override agent behavior. * Ingestion points: The skill uses read_resource and search tools to retrieve table descriptions, field descriptions, and glossary terms from the Metabase server (SKILL.md). * Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when processing retrieved metadata. * Capability inventory: The agent can execute SQL queries, create Metabase entities, and perform file-write operations. * Sanitization: There is no evidence of sanitization or escaping of the metadata retrieved from the database before it is used in prompts or logic.
  • [DYNAMIC_EXECUTION]: The skill facilitates the creation and execution of SQL queries and Metabase 'Transforms' to verify data logic and outcomes. These operations are performed via the execute_query and execute_sql tools provided by the Metabase MCP server.
  • [DATA_EXPOSURE]: The skill writes session progress, including user roles and plan details, to a local file at ./.claude/ai-readiness-checklist/progress.json.
  • [SAFE]: External links to metabase.com are vendor-owned resources for documentation and setup.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 03:10 AM
Security Audit — agent-trust-hub — ai-readiness-checklist