Update to Zid | زد

{
  "description": "Connect and automate a Zid store. Use to connect an AI agent, n8n, Make or a script to the Zid MCP server, schedule a recurring store job, or fix a Zid MCP tool error in a saved workflow.",
  "included_files": [
    {
      "relative_path": "references/jobs.md",
      "size_in_bytes": 2940
    },
    {
      "relative_path": "references/mcp.md",
      "size_in_bytes": 3646
    }
  ],
  "name": "zid-ops",
  "skill_md_contents": "---\nname: zid-ops\ndescription: Connect and automate a Zid store. Use to connect an AI agent, n8n, Make or a script to the Zid MCP server, schedule a recurring store job, or fix a Zid MCP tool error in a saved workflow.\n---\n\n# Zid ops\n\nRunning a Zid store with automations: connecting clients to the Zid MCP server\nand scheduling store jobs. The other Zid skills describe the store work; this\none covers how to run it without a chat. Building an app on the Merchant API,\nOAuth or webhooks is the `zid-api` skill.\n\n## Ground rules\n\n- The user's explicit instructions take precedence over this skill.\n- Secrets stay out of chat: the personal MCP URL and any tool credentials go\n  straight into the client's or automation tool's credential settings. In\n  examples, write them as placeholders such as `<MCP_URL>`.\n- One writer per field. Before automating a write, check whether an installed app\n  already writes it (`zid_account.list_installed_apps`) and agree with the\n  merchant which one owns it.\n\n## Pick the topic\n\n| Topic | Use it when | Read |\n| --- | --- | --- |\n| MCP connections | Connecting Claude, Cursor, Codex, n8n, Make or a script to the store's tools; tool errors and statuses | [references/mcp.md](references/mcp.md) |\n| Scheduled jobs | A store task should run every day or on an event, without someone in the chat | [references/jobs.md](references/jobs.md) |\n\nRead the topic's reference before writing anything for it; a job usually needs both.\n\n## Steps\n\n1. Pick the topic and read its reference.\n2. Settle the inputs: which store, which client or tool, what triggers the run,\n   what it reads and what it may change.\n3. Produce the artifact: client configuration or workflow steps, with secrets as placeholders.\n4. Prove it with a read before any write: call one named read action, such as\n   `zid_store.get_context` or `zid_orders.list_orders`.\n5. Hand over: where each secret goes, how to rotate it, and how to tell when the\n   automation fails.\n\nDone when the client or job has run one read successfully and the merchant knows\nwhere its secret lives.\n\n## Common failures\n\nAnswers such as `invalid_arguments`, `confirmation_required`, the stale-plan\nstatuses and `429` are covered in [references/mcp.md](references/mcp.md), under\n\"How the tools behave\".\n\n| Symptom | Cause | Fix |\n| --- | --- | --- |\n| Unknown tool in a saved workflow | The workflow uses a renamed tool | Replace the step with the tool and action the matching Zid skill names |\n| A skill's tool is not available | The URL loads only some areas | Add the area to `?toolsets=` ([references/mcp.md](references/mcp.md)) |\n| The URL stopped working | It was rotated | Copy the new URL from the Zid AI Connector app into every client |\n"
}