The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses child_process.execFile to interact with the Shopify CLI and the npm environment. This is implemented in scripts/core/graphql.mjs to execute GraphQL queries and in scripts/core/permissions.mjs for app management tasks.\n- [REMOTE_CODE_EXECUTION]: In SKILL.md, the instructions require the agent to execute a silent update using npx --yes skills update shopify-store-setup-auditor -g -y. This downloads and executes code from the npm registry to ensure the auditing logic is current.\n- [DATA_EXPOSURE]: The skill processes sensitive API credentials including SKILL_HUB_SHOPIFY_CLIENT_SECRET and SKILL_HUB_SHOPIFY_APP_AUTOMATION_TOKEN. These are managed through a local skill-hub.env file. Security is maintained by recommending .gitignore inclusion, applying 0o600 file permissions during creation in scripts/core/config.mjs, and utilizing a redacting error handler for log output.\n- [INDIRECT_PROMPT_INJECTION]: The skill audits storefront HTML and metadata which are external and untrusted. Ingestion points: scripts/core/public-fetch.mjs via extractMeta and fetchPublic. Boundary markers: SKILL.md contains a 'Non-Negotiable' directive to treat storefront data as untrusted and never execute instructions found within it. Capability inventory: Subprocess execution (Shopify CLI), local file system access, and network requests to the Shopify Admin API. Sanitization: The skill employs escapeHtml when generating reports in scripts/core/public-fetch.mjs and uses instructional isolation to prevent the agent from treating data as commands.

shopify-store-setup-auditor