[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F3PQWMADW75VJD9MXXVAMG":3,"skill-same:01M7F3PQWMADW75VJD9MXXVAMG":63},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":29,"updatedAt":30,"install":31,"resources":44},"01M7F3PQWMADW75VJD9MXXVAMG","n8n-validation-expert",null,"Interpret validation errors and guide fixing them. Use when encountering validation errors, validation warnings, false positives, operator structure issues, or need help understanding validation results. Also use when asking about validation profiles, error types, the validation loop process, or auto-fix capabilities. Consult this skill whenever a validate_node or validate_workflow call returns errors or warnings — it knows which warnings are false positives and which errors need real fixes.","26BB\u002Fagent-fleet-os","26BB","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F113676969?v=4","01M7F3AVX9MTTAHCYTANRGEE52","Bhushan Vijay Bhosale","01M7F3AVY15DJBX1CE4H25P52T",0,"MIT","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002FN8N agent\u002Fplaybooks\u002Fn8n-validation-expert\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: n8n-validation-expert\ndescription: Interpret validation errors and guide fixing them. Use when encountering validation errors, validation warnings, false positives, operator structure issues, or need help understanding validation results. Also use when asking about validation profiles, error types, the validation loop process, or auto-fix capabilities. Consult this skill whenever a validate_node or validate_workflow call returns errors or warnings — it knows which warnings are false positives and which errors need real fixes.\n---\n\n# n8n Validation Expert\n\nExpert guide for interpreting and fixing n8n validation errors.\n\n---\n\n## Validation Philosophy\n\n**Validate early, validate often**\n\nValidation is typically iterative:\n- Expect validation feedback loops\n- Usually 2-3 validate → fix cycles\n- Average: 23s thinking about errors, 58s fixing them\n\n**Key insight**: Validation is an iterative process, not one-shot!\n\n---\n\n## Error Severity Levels\n\n### 1. Errors (Must Fix)\n**Blocks workflow execution** - Must be resolved before activation\n\n**Types**:\n- `missing_required` - Required field not provided\n- `invalid_value` - Value doesn't match allowed options\n- `type_mismatch` - Wrong data type (string instead of number)\n- `invalid_reference` - Referenced node doesn't exist\n- `invalid_expression` - Expression syntax error\n\n**Example**:\n```json\n{\n  \"type\": \"missing_required\",\n  \"property\": \"channel\",\n  \"message\": \"Channel name is required\",\n  \"fix\": \"Provide a channel name (lowercase, no spaces, 1-80 characters)\"\n}\n```\n\n### 2. Warnings (Should Fix)\n**Doesn't block execution** - Workflow can be activated but may have issues\n\n**Types**:\n- `best_practice` - Recommended but not required — surfaces under `ai-friendly` \u002F `strict` only\n- `deprecated` - Using old API\u002Ffeature — surfaces under every profile\n- `security` - Hardcoded secrets, unauthenticated webhooks — surfaces under every profile\n- `performance` - Potential performance issue — advisory, `ai-friendly` \u002F `strict`\n\n**Example** (best-practice — appears under `ai-friendly` \u002F `strict`):\n```json\n{\n  \"type\": \"warning\",\n  \"nodeName\": \"Slack\",\n  \"message\": \"Slack API can have rate limits and transient failures\"\n}\n```\n\n### 3. Suggestions (Optional)\n**Nice to have** - Improvements that could enhance workflow\n\n**Types**:\n- `optimization` - Could be more efficient\n- `alternative` - Better way to achieve same result\n\n---\n\n## The Validation Loop\n\n### Pattern from Telemetry\n**7,841 occurrences** of this pattern:\n\n```\n1. Configure node\n   ↓\n2. validate_node (23 seconds thinking about errors)\n   ↓\n3. Read error messages carefully\n   ↓\n4. Fix errors\n   ↓\n5. validate_node again (58 seconds fixing)\n   ↓\n6. Repeat until valid (usually 2-3 iterations)\n```\n\n### Example\n```javascript\n\u002F\u002F Iteration 1\nlet config = {\n  resource: \"channel\",\n  operation: \"create\"\n};\n\nconst result1 = validate_node({\n  nodeType: \"nodes-base.slack\",\n  config,\n  profile: \"runtime\"\n});\n\u002F\u002F → Error: Missing \"name\"\n\n\u002F\u002F ⏱️  23 seconds thinking...\n\n\u002F\u002F Iteration 2\nconfig.name = \"general\";\n\nconst result2 = validate_node({\n  nodeType: \"nodes-base.slack\",\n  config,\n  profile: \"runtime\"\n});\n\u002F\u002F → Error: Missing \"text\"\n\n\u002F\u002F ⏱️  58 seconds fixing...\n\n\u002F\u002F Iteration 3\nconfig.text = \"Hello!\";\n\nconst result3 = validate_node({\n  nodeType: \"nodes-base.slack\",\n  config,\n  profile: \"runtime\"\n});\n\u002F\u002F → Valid! ✅\n```\n\n**This is normal!** Don't be discouraged by multiple iterations.\n\n---\n\n## Validation Profiles\n\nThe four profiles are **cumulative** (n8n-mcp ≥ 2.63.0): each surfaces everything the lower one does, plus more. The dividing line is best-practice *advisories* — `minimal` and `runtime` withhold them; `ai-friendly` and `strict` add them. Errors are the same across every profile except that `minimal` skips a few config-level checks (e.g. enum validation of an explicit `operation`). Security and deprecation warnings surface under every profile.\n\n### minimal\n**Use when**: Quick structural checks while wiring a workflow together.\n\n**Surfaces**: hard errors that would stop execution (missing required fields, empty code, broken connections). Skips enum checks and all advisories.\n\n**Fastest and most permissive.**\n\n### runtime (RECOMMENDED default)\n**Use when**: Ongoing validation as you build; the everyday profile.\n\n**Surfaces**: errors (required fields, value types, allowed values, dependencies, broken references) plus security and deprecation warnings. **No** best-practice advisories.\n\n**Balanced — catches everything that breaks, stays quiet about style.**\n\n### ai-friendly\n**Use when**: You want the best-practice advice before deploying.\n\n**Surfaces**: everything `runtime` does, **plus** best-practice advisories — per-node \"without error handling\" suggestions, \"webhook should always send a response\", rate-limit notes, outdated-`typeVersion` suggestions, `cachedResultName` and long-chain hints.\n\n**Note**: `ai-friendly` is *stricter* than `runtime`, not looser. (Older docs described it as reducing false positives — that was true only while profile gating was broken; it is fixed now.)\n\n### strict\n**Use when**: Hardening a production-critical workflow.\n\n**Surfaces**: everything `ai-friendly` does, **plus** leftover-property checks (\"property 'X' won't be used — not visible with current settings\").\n\n**Maximum lint.** With the false positives fixed at the source, its warnings are advice to weigh, not noise to fight.\n\n---\n\n## Common Error Types\n\nFive core error types, in rough order of frequency:\n\n- **`missing_required`** — a required field isn't provided. Use `get_node` to see required fields, then add it.\n- **`invalid_value`** — value doesn't match allowed options (enums are case-sensitive). Check the error's allowed list or `get_node`.\n- **`type_mismatch`** — wrong data type (string `\"100\"` vs number `100`). Convert to the expected type.\n- **`invalid_expression`** — expression syntax error (missing `{{}}`, typos). See the n8n Expression Syntax skill.\n- **`invalid_reference`** — referenced node doesn't exist (renamed, deleted, or misspelled). Fix the name or `cleanStaleConnections`.\n\nA sixth class, **`patchNodeField` errors** (find-not-found, ambiguous match, invalid\u002Funsafe regex), surfaces when a `patchNodeField` op fails during `n8n_update_partial_workflow` — it's strict by design and errors rather than silently continuing.\n\nEvery type above has worked examples (broken config → fix) plus the patchNodeField error cases and their fixes in **[ERROR_CATALOG.md](ERROR_CATALOG.md)**.\n\n---\n\n## Auto-Sanitization System\n\n**Automatically normalizes common operator structures** on ANY workflow update — `n8n_create_workflow`, `n8n_update_partial_workflow`, or any save. Trust it; don't hand-fix these.\n\n**What it normalizes on save**:\n- **Binary operators** (equals, notEquals, contains, notContains, greaterThan, lessThan, startsWith, endsWith) — removes a stray `singleValue` property.\n- **Unary operators** (isEmpty, isNotEmpty, true, false) — adds `singleValue: true`.\n- **IF\u002FSwitch metadata** — fills in `conditions.options` for IF v2.2+ and Switch v3.2+.\n\n**Validation no longer errors on these shapes** (n8n-mcp ≥ 2.63.0). n8n derives unary-ness from the operator name and defaults the `conditions.options` sub-fields, so `validate_node` \u002F `validate_workflow` accept a condition whether or not `singleValue` and the options metadata are present — the sanitizer just tidies the canonical form on save. (Older servers wrongly errored on the un-normalized shape; if you see that, upgrade.) What still *is* a real error: a v1-shaped `conditions` object on a v2 node, an empty filter with no conditions, and legacy v1 operator names (e.g. `smaller`) inside a v2 structure.\n\n**What the sanitizer CANNOT fix** (handle manually): broken connections to non-existent nodes (use `cleanStaleConnections`), branch-count mismatches (add\u002Fremove connections or rules), and paradoxical corrupt states (may need manual DB intervention).\n\nBefore\u002Fafter examples and the full cannot-fix detail are in **[ERROR_CATALOG.md](ERROR_CATALOG.md)** (Auto-Sanitization sections).\n\n---\n\n## False Positives\n\nThe validator overhaul (n8n-mcp ≥ 2.63.0) removed the classic false positives — template literals inside expressions, optional chaining, omitted-operation defaults, the Webhook → Respond-to-Webhook pattern, IF\u002FFilter legacy shapes, and more no longer fire. There is no standing list of \"known false positives to ignore.\"\n\nWhat remains are **best-practice advisories** (surfaced only under `ai-friendly` \u002F `strict`) that flag a real trade-off but may be acceptable in your case. Not every advisory needs a fix — many are context-dependent. Common ones and when each is acceptable vs. worth fixing:\n\n- **\"...without error handling\"** — OK for dev\u002Ftesting and non-critical notifications; fix for production handling important data. (Never a hard error — style doesn't block execution.)\n- **\"No retry logic\"** — OK for idempotent ops, APIs with their own retry, manual triggers; fix for flaky external services and production automation.\n- **\"...rate limits and transient failures\"** — OK for internal\u002Flow-volume\u002Fserver-side-limited APIs; fix for public, high-volume APIs.\n- **\"Unbounded query\"** — OK for small known datasets, aggregations, dev\u002Ftesting; fix for production queries on large tables.\n\nSecurity and deprecation warnings, by contrast, surface under *every* profile and should be treated as real.\n\nFull per-case guidance, the list of what the validator no longer flags, profile strategies, the \"should I fix this?\" decision framework, and how to document accepted advisories are in **[FALSE_POSITIVES.md](FALSE_POSITIVES.md)**.\n\n---\n\n## Validation Result Structure\n\n### Complete Response\n```javascript\n{\n  \"valid\": false,\n  \"errors\": [\n    {\n      \"type\": \"missing_required\",\n      \"property\": \"channel\",\n      \"message\": \"Channel name is required\",\n      \"fix\": \"Provide a channel name (lowercase, no spaces)\"\n    }\n  ],\n  \"warnings\": [\n    {\n      \"type\": \"best_practice\",\n      \"property\": \"errorHandling\",\n      \"message\": \"Slack API can have rate limits\",\n      \"suggestion\": \"Add onError: 'continueRegularOutput'\"\n    }\n  ],\n  \"suggestions\": [\n    {\n      \"type\": \"optimization\",\n      \"message\": \"Consider using batch operations for multiple messages\"\n    }\n  ],\n  \"summary\": {\n    \"hasErrors\": true,\n    \"errorCount\": 1,\n    \"warningCount\": 1,\n    \"suggestionCount\": 1\n  }\n}\n```\n\n### How to Read It\n\n1. **Check `valid` first** — `true` means the config is valid; `false` means there are errors to fix before deployment.\n2. **Fix `errors` first** — each carries a `property`, `message`, and `fix`. These must be resolved.\n3. **Review `warnings`** — each has a `message` and `suggestion`; decide per-case whether to address it (see False Positives above).\n4. **Consider `suggestions`** — optional improvements, not required.\n\n---\n\n## Workflow Validation\n\n### validate_workflow (Structure)\n**Validates entire workflow**, not just individual nodes\n\n**Checks**:\n1. **Node configurations** - Each node valid\n2. **Connections** - No broken references\n3. **Expressions** - Syntax and references valid\n4. **Flow** - Logical workflow structure\n\n**Example**:\n```javascript\nvalidate_workflow({\n  workflow: {\n    nodes: [...],\n    connections: {...}\n  },\n  options: {\n    validateNodes: true,\n    validateConnections: true,\n    validateExpressions: true,\n    profile: \"runtime\"\n  }\n})\n```\n\n### Common Workflow Errors\n\n#### 1. Broken Connections\n```json\n{\n  \"error\": \"Connection from 'Transform' to 'NonExistent' - target node not found\"\n}\n```\n\n**Fix**: Remove stale connection or create missing node\n\n#### 2. Cycles (warning, not an error)\n```json\n{\n  \"warning\": \"Workflow contains a cycle: Node A → Node B → Node A\"\n}\n```\n\nA cycle is a **warning**, not a hard error (n8n-mcp ≥ 2.63.0) — runtime-controlled loops (error-retry, data-driven pagination, a router feeding back) execute to completion and are legitimate. **Fix** only if the loop is unintentional: ensure the cycle has a real exit (a conditional node, an error output, or a bounded counter) so it can't spin forever.\n\n#### 3. Multiple Start Nodes\n```json\n{\n  \"warning\": \"Multiple trigger nodes found - only one will execute\"\n}\n```\n\n**Fix**: Remove extra triggers or split into separate workflows\n\n#### 4. Disconnected Nodes\n```json\n{\n  \"warning\": \"Node 'Transform' is not connected to workflow flow\"\n}\n```\n\n**Fix**: Connect node or remove if unused\n\n---\n\n## Recovery Strategies\n\n### Strategy 1: Start Fresh\n**When**: Configuration is severely broken\n\n**Steps**:\n1. Note required fields from `get_node`\n2. Create minimal valid configuration\n3. Add features incrementally\n4. Validate after each addition\n\n### Strategy 2: Binary Search\n**When**: Workflow validates but executes incorrectly\n\n**Steps**:\n1. Remove half the nodes\n2. Validate and test\n3. If works: problem is in removed nodes\n4. If fails: problem is in remaining nodes\n5. Repeat until problem isolated\n\n### Strategy 3: Clean Stale Connections\n**When**: \"Node not found\" errors\n\n**Steps**:\n```javascript\nn8n_update_partial_workflow({\n  id: \"workflow-id\",\n  operations: [{\n    type: \"cleanStaleConnections\"\n  }]\n})\n```\n\n### Strategy 4: Use Auto-fix\n**When**: Validation errors that can be automatically resolved\n\n**Steps**:\n```javascript\n\u002F\u002F Preview fixes (default - doesn't apply)\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  applyFixes: false,\n  confidenceThreshold: \"medium\"  \u002F\u002F high, medium, low\n})\n\n\u002F\u002F Review fixes, then apply\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  applyFixes: true\n})\n```\n\n---\n\n## Auto-Fix Capabilities\n\nThe `n8n_autofix_workflow` tool can fix these issue types:\n\n1. **expression-format** - Missing `=` prefix in expressions (e.g., `{{ $json.field }}` → `={{ $json.field }}`)\n2. **typeversion-correction** - Downgrades nodes with unsupported typeVersions\n3. **error-output-config** - Removes conflicting onError settings\n4. **node-type-correction** - Fixes unknown node types using similarity matching (90%+ confidence)\n5. **webhook-missing-path** - Generates UUIDs for webhook nodes missing path configuration\n6. **typeversion-upgrade** - Smart upgrades to latest node versions with auto-migration\n7. **version-migration** - Guidance for complex breaking changes requiring manual steps\n\n**Confidence levels**: `high` (90%+, safe to auto-apply), `medium` (70-89%, review recommended), `low` (\u003C70%, manual review required)\n\n```javascript\n\u002F\u002F Preview all fixes\nn8n_autofix_workflow({id: \"workflow-id\"})\n\n\u002F\u002F Only apply high-confidence fixes\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  applyFixes: true,\n  confidenceThreshold: \"high\"\n})\n\n\u002F\u002F Target specific fix types\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  fixTypes: [\"expression-format\", \"typeversion-upgrade\"],\n  applyFixes: true\n})\n```\n\n**Post-update guidance**: For version upgrades, check the `postUpdateGuidance` field in the response for step-by-step migration instructions.\n\n---\n\n## Best Practices\n\n### ✅ Do\n\n- Validate after every significant change\n- Read error messages completely\n- Fix errors iteratively (one at a time)\n- Use `runtime` profile for pre-deployment\n- Check `valid` field before assuming success\n- Trust auto-sanitization for operator issues\n- Use `get_node` when unclear about requirements\n- Document false positives you accept\n\n### ❌ Don't\n\n- Skip validation before activation\n- Try to fix all errors at once\n- Ignore error messages\n- Use `strict` profile during development (too noisy)\n- Assume validation passed (always check result)\n- Manually fix auto-sanitization issues\n- Deploy with unresolved errors\n- Ignore all warnings (some are important!)\n\n---\n\n## Running the workflow after it validates\n\n`validate_workflow` checks structure, parameters and expressions — it never runs anything. A workflow that validates cleanly can still fail on real data, so run it once before calling it done.\n\n**With a webhook, form or chat trigger:** `n8n_test_workflow({workflowId})` — the default `method: \"auto\"` detects the trigger and fires it over HTTP (the workflow must be active).\n\n**Without such a trigger** (Manual Trigger, Schedule, sub-workflow) there is no HTTP entry point. Use the pin-data path, which runs through n8n's own MCP server (`N8N_MCP_ACCESS_TOKEN`, n8n 2.34+):\n\n1. `n8n_test_workflow({workflowId, method: \"prepare\"})` — lists the nodes that need pinned data.\n2. Build one sample item per listed node, keyed by node **name**, each item wrapped in `{json: {...}}`:\n   ```json\n   {\"When clicking 'Test workflow'\": [{\"json\": {\"orderId\": \"1234\", \"email\": \"a@b.com\"}}]}\n   ```\n   A flat object instead of an array of `{json}` items is the usual mistake here.\n3. `n8n_test_workflow({workflowId, method: \"pinned\", pinData})` — runs it with that data and waits for the result.\n\n**For a quick manual run without pinned data**, `method: \"direct\"` starts a manual execution and returns as soon as it has started; poll `n8n_executions({action: \"get\", id: executionId, mode: \"error\"})` for the outcome. Nothing is pinned on a `direct` run, so every node executes and any external call it makes is real — and even `pinned` only pins trigger, credentialed and HTTP Request nodes. `executionMode: \"production\"` changes the execution context, not whether there are side effects; leave it at the default unless the user asked for a production run.\n\n**`method: \"auto\"` never runs a workflow through n8n's MCP server.** On a workflow with no external trigger it reports that fact and names `prepare`\u002F`pinned`\u002F`direct`; the routed methods only run when you ask for them by name.\n\n**Consent before the first routed run.** n8n refuses these calls for a workflow whose \"Available in MCP\" setting is off, which comes back as `WORKFLOW_NOT_EXPOSED`. Re-running with `exposeToMcp: true` turns that setting on and retries once. It is a visible, persistent setting on the workflow (and enabling it is a workflow update, so a concurrent UI edit can be overwritten) — **ask the user before passing it**. The consent flow only ever enables the setting; nothing disables it implicitly. Turning it off again is a deliberate act: `n8n_update_partial_workflow({id: workflowId, operations: [{type: \"updateSettings\", settings: {availableInMCP: false}}]})`, or the toggle in the n8n UI.\n\n**Reading the result:** a run that started and then failed comes back as `EXECUTION_FAILED` with the `executionId` — inspect it with `n8n_executions({action: \"get\", id, mode: \"error\"})` and fix from the node that threw, then validate and run again.\n\n---\n\n## Reviewing an existing workflow\n\nValidating as you build (the loop above) is for catching schema and shape errors in your own in-progress work. **Reviewing an existing workflow** — yours or one you've been handed — is a different job: the workflow already passes `validate_workflow` clean, and you're hunting for the issues validation doesn't see (silent connection bugs, injection-prone queries, dropped-item Switches, Set\u002FCode antipatterns, missing error paths). For that, pull the workflow with `n8n_get_workflow` and walk **[REVIEW_CHECKLIST.md](REVIEW_CHECKLIST.md)** — a severity-tiered audit (MUST FIX \u002F SHOULD FIX \u002F NICE TO HAVE) where every item points to the canonical skill for the fix. Run `n8n_audit_instance` alongside it to surface hardcoded secrets and unauthenticated webhooks across the whole instance.\n\n---\n\n## Detailed Guides\n\nFor comprehensive error catalogs, false positives, and workflow review:\n\n- **[ERROR_CATALOG.md](ERROR_CATALOG.md)** - Complete list of error types with examples\n- **[FALSE_POSITIVES.md](FALSE_POSITIVES.md)** - When warnings are acceptable\n- **[REVIEW_CHECKLIST.md](REVIEW_CHECKLIST.md)** - Severity-tiered audit for reviewing an existing workflow\n\n---\n\n## Summary\n\n**Key Points**:\n1. **Validation is iterative** (avg 2-3 cycles, 23s + 58s)\n2. **Errors must be fixed**, warnings are optional\n3. **Auto-sanitization** normalizes operator structures on save; validation no longer errors on the raw shape\n4. **Use runtime profile** by default; step up to `ai-friendly`\u002F`strict` for best-practice advisories\n5. **Classic false positives are fixed** (≥ 2.63.0) — remaining warnings are advisories or security\u002Fdeprecation notices, not validator mistakes\n6. **Read error messages** - they contain fix guidance\n\n**Validation Process**:\n1. Validate → Read errors → Fix → Validate again\n2. Repeat until valid (usually 2-3 iterations)\n3. Review warnings and decide if acceptable\n4. Deploy with confidence\n\n**Related Skills & Tools**:\n- n8n MCP Tools Expert - Use validation tools correctly\n- n8n Expression Syntax - Fix expression errors\n- n8n Node Configuration - Understand required fields\n- `n8n_audit_instance` - Proactive security validation (hardcoded secrets, unauthenticated webhooks, missing error handling, data retention)\n","# n8n Validation Expert\n\nExpert guide for interpreting and fixing n8n validation errors.\n\n---\n\n## Validation Philosophy\n\n**Validate early, validate often**\n\nValidation is typically iterative:\n- Expect validation feedback loops\n- Usually 2-3 validate → fix cycles\n- Average: 23s thinking about errors, 58s fixing them\n\n**Key insight**: Validation is an iterative process, not one-shot!\n\n---\n\n## Error Severity Levels\n\n### 1. Errors (Must Fix)\n**Blocks workflow execution** - Must be resolved before activation\n\n**Types**:\n- `missing_required` - Required field not provided\n- `invalid_value` - Value doesn't match allowed options\n- `type_mismatch` - Wrong data type (string instead of number)\n- `invalid_reference` - Referenced node doesn't exist\n- `invalid_expression` - Expression syntax error\n\n**Example**:\n```json\n{\n  \"type\": \"missing_required\",\n  \"property\": \"channel\",\n  \"message\": \"Channel name is required\",\n  \"fix\": \"Provide a channel name (lowercase, no spaces, 1-80 characters)\"\n}\n```\n\n### 2. Warnings (Should Fix)\n**Doesn't block execution** - Workflow can be activated but may have issues\n\n**Types**:\n- `best_practice` - Recommended but not required — surfaces under `ai-friendly` \u002F `strict` only\n- `deprecated` - Using old API\u002Ffeature — surfaces under every profile\n- `security` - Hardcoded secrets, unauthenticated webhooks — surfaces under every profile\n- `performance` - Potential performance issue — advisory, `ai-friendly` \u002F `strict`\n\n**Example** (best-practice — appears under `ai-friendly` \u002F `strict`):\n```json\n{\n  \"type\": \"warning\",\n  \"nodeName\": \"Slack\",\n  \"message\": \"Slack API can have rate limits and transient failures\"\n}\n```\n\n### 3. Suggestions (Optional)\n**Nice to have** - Improvements that could enhance workflow\n\n**Types**:\n- `optimization` - Could be more efficient\n- `alternative` - Better way to achieve same result\n\n---\n\n## The Validation Loop\n\n### Pattern from Telemetry\n**7,841 occurrences** of this pattern:\n\n```\n1. Configure node\n   ↓\n2. validate_node (23 seconds thinking about errors)\n   ↓\n3. Read error messages carefully\n   ↓\n4. Fix errors\n   ↓\n5. validate_node again (58 seconds fixing)\n   ↓\n6. Repeat until valid (usually 2-3 iterations)\n```\n\n### Example\n```javascript\n\u002F\u002F Iteration 1\nlet config = {\n  resource: \"channel\",\n  operation: \"create\"\n};\n\nconst result1 = validate_node({\n  nodeType: \"nodes-base.slack\",\n  config,\n  profile: \"runtime\"\n});\n\u002F\u002F → Error: Missing \"name\"\n\n\u002F\u002F ⏱️  23 seconds thinking...\n\n\u002F\u002F Iteration 2\nconfig.name = \"general\";\n\nconst result2 = validate_node({\n  nodeType: \"nodes-base.slack\",\n  config,\n  profile: \"runtime\"\n});\n\u002F\u002F → Error: Missing \"text\"\n\n\u002F\u002F ⏱️  58 seconds fixing...\n\n\u002F\u002F Iteration 3\nconfig.text = \"Hello!\";\n\nconst result3 = validate_node({\n  nodeType: \"nodes-base.slack\",\n  config,\n  profile: \"runtime\"\n});\n\u002F\u002F → Valid! ✅\n```\n\n**This is normal!** Don't be discouraged by multiple iterations.\n\n---\n\n## Validation Profiles\n\nThe four profiles are **cumulative** (n8n-mcp ≥ 2.63.0): each surfaces everything the lower one does, plus more. The dividing line is best-practice *advisories* — `minimal` and `runtime` withhold them; `ai-friendly` and `strict` add them. Errors are the same across every profile except that `minimal` skips a few config-level checks (e.g. enum validation of an explicit `operation`). Security and deprecation warnings surface under every profile.\n\n### minimal\n**Use when**: Quick structural checks while wiring a workflow together.\n\n**Surfaces**: hard errors that would stop execution (missing required fields, empty code, broken connections). Skips enum checks and all advisories.\n\n**Fastest and most permissive.**\n\n### runtime (RECOMMENDED default)\n**Use when**: Ongoing validation as you build; the everyday profile.\n\n**Surfaces**: errors (required fields, value types, allowed values, dependencies, broken references) plus security and deprecation warnings. **No** best-practice advisories.\n\n**Balanced — catches everything that breaks, stays quiet about style.**\n\n### ai-friendly\n**Use when**: You want the best-practice advice before deploying.\n\n**Surfaces**: everything `runtime` does, **plus** best-practice advisories — per-node \"without error handling\" suggestions, \"webhook should always send a response\", rate-limit notes, outdated-`typeVersion` suggestions, `cachedResultName` and long-chain hints.\n\n**Note**: `ai-friendly` is *stricter* than `runtime`, not looser. (Older docs described it as reducing false positives — that was true only while profile gating was broken; it is fixed now.)\n\n### strict\n**Use when**: Hardening a production-critical workflow.\n\n**Surfaces**: everything `ai-friendly` does, **plus** leftover-property checks (\"property 'X' won't be used — not visible with current settings\").\n\n**Maximum lint.** With the false positives fixed at the source, its warnings are advice to weigh, not noise to fight.\n\n---\n\n## Common Error Types\n\nFive core error types, in rough order of frequency:\n\n- **`missing_required`** — a required field isn't provided. Use `get_node` to see required fields, then add it.\n- **`invalid_value`** — value doesn't match allowed options (enums are case-sensitive). Check the error's allowed list or `get_node`.\n- **`type_mismatch`** — wrong data type (string `\"100\"` vs number `100`). Convert to the expected type.\n- **`invalid_expression`** — expression syntax error (missing `{{}}`, typos). See the n8n Expression Syntax skill.\n- **`invalid_reference`** — referenced node doesn't exist (renamed, deleted, or misspelled). Fix the name or `cleanStaleConnections`.\n\nA sixth class, **`patchNodeField` errors** (find-not-found, ambiguous match, invalid\u002Funsafe regex), surfaces when a `patchNodeField` op fails during `n8n_update_partial_workflow` — it's strict by design and errors rather than silently continuing.\n\nEvery type above has worked examples (broken config → fix) plus the patchNodeField error cases and their fixes in **[ERROR_CATALOG.md](ERROR_CATALOG.md)**.\n\n---\n\n## Auto-Sanitization System\n\n**Automatically normalizes common operator structures** on ANY workflow update — `n8n_create_workflow`, `n8n_update_partial_workflow`, or any save. Trust it; don't hand-fix these.\n\n**What it normalizes on save**:\n- **Binary operators** (equals, notEquals, contains, notContains, greaterThan, lessThan, startsWith, endsWith) — removes a stray `singleValue` property.\n- **Unary operators** (isEmpty, isNotEmpty, true, false) — adds `singleValue: true`.\n- **IF\u002FSwitch metadata** — fills in `conditions.options` for IF v2.2+ and Switch v3.2+.\n\n**Validation no longer errors on these shapes** (n8n-mcp ≥ 2.63.0). n8n derives unary-ness from the operator name and defaults the `conditions.options` sub-fields, so `validate_node` \u002F `validate_workflow` accept a condition whether or not `singleValue` and the options metadata are present — the sanitizer just tidies the canonical form on save. (Older servers wrongly errored on the un-normalized shape; if you see that, upgrade.) What still *is* a real error: a v1-shaped `conditions` object on a v2 node, an empty filter with no conditions, and legacy v1 operator names (e.g. `smaller`) inside a v2 structure.\n\n**What the sanitizer CANNOT fix** (handle manually): broken connections to non-existent nodes (use `cleanStaleConnections`), branch-count mismatches (add\u002Fremove connections or rules), and paradoxical corrupt states (may need manual DB intervention).\n\nBefore\u002Fafter examples and the full cannot-fix detail are in **[ERROR_CATALOG.md](ERROR_CATALOG.md)** (Auto-Sanitization sections).\n\n---\n\n## False Positives\n\nThe validator overhaul (n8n-mcp ≥ 2.63.0) removed the classic false positives — template literals inside expressions, optional chaining, omitted-operation defaults, the Webhook → Respond-to-Webhook pattern, IF\u002FFilter legacy shapes, and more no longer fire. There is no standing list of \"known false positives to ignore.\"\n\nWhat remains are **best-practice advisories** (surfaced only under `ai-friendly` \u002F `strict`) that flag a real trade-off but may be acceptable in your case. Not every advisory needs a fix — many are context-dependent. Common ones and when each is acceptable vs. worth fixing:\n\n- **\"...without error handling\"** — OK for dev\u002Ftesting and non-critical notifications; fix for production handling important data. (Never a hard error — style doesn't block execution.)\n- **\"No retry logic\"** — OK for idempotent ops, APIs with their own retry, manual triggers; fix for flaky external services and production automation.\n- **\"...rate limits and transient failures\"** — OK for internal\u002Flow-volume\u002Fserver-side-limited APIs; fix for public, high-volume APIs.\n- **\"Unbounded query\"** — OK for small known datasets, aggregations, dev\u002Ftesting; fix for production queries on large tables.\n\nSecurity and deprecation warnings, by contrast, surface under *every* profile and should be treated as real.\n\nFull per-case guidance, the list of what the validator no longer flags, profile strategies, the \"should I fix this?\" decision framework, and how to document accepted advisories are in **[FALSE_POSITIVES.md](FALSE_POSITIVES.md)**.\n\n---\n\n## Validation Result Structure\n\n### Complete Response\n```javascript\n{\n  \"valid\": false,\n  \"errors\": [\n    {\n      \"type\": \"missing_required\",\n      \"property\": \"channel\",\n      \"message\": \"Channel name is required\",\n      \"fix\": \"Provide a channel name (lowercase, no spaces)\"\n    }\n  ],\n  \"warnings\": [\n    {\n      \"type\": \"best_practice\",\n      \"property\": \"errorHandling\",\n      \"message\": \"Slack API can have rate limits\",\n      \"suggestion\": \"Add onError: 'continueRegularOutput'\"\n    }\n  ],\n  \"suggestions\": [\n    {\n      \"type\": \"optimization\",\n      \"message\": \"Consider using batch operations for multiple messages\"\n    }\n  ],\n  \"summary\": {\n    \"hasErrors\": true,\n    \"errorCount\": 1,\n    \"warningCount\": 1,\n    \"suggestionCount\": 1\n  }\n}\n```\n\n### How to Read It\n\n1. **Check `valid` first** — `true` means the config is valid; `false` means there are errors to fix before deployment.\n2. **Fix `errors` first** — each carries a `property`, `message`, and `fix`. These must be resolved.\n3. **Review `warnings`** — each has a `message` and `suggestion`; decide per-case whether to address it (see False Positives above).\n4. **Consider `suggestions`** — optional improvements, not required.\n\n---\n\n## Workflow Validation\n\n### validate_workflow (Structure)\n**Validates entire workflow**, not just individual nodes\n\n**Checks**:\n1. **Node configurations** - Each node valid\n2. **Connections** - No broken references\n3. **Expressions** - Syntax and references valid\n4. **Flow** - Logical workflow structure\n\n**Example**:\n```javascript\nvalidate_workflow({\n  workflow: {\n    nodes: [...],\n    connections: {...}\n  },\n  options: {\n    validateNodes: true,\n    validateConnections: true,\n    validateExpressions: true,\n    profile: \"runtime\"\n  }\n})\n```\n\n### Common Workflow Errors\n\n#### 1. Broken Connections\n```json\n{\n  \"error\": \"Connection from 'Transform' to 'NonExistent' - target node not found\"\n}\n```\n\n**Fix**: Remove stale connection or create missing node\n\n#### 2. Cycles (warning, not an error)\n```json\n{\n  \"warning\": \"Workflow contains a cycle: Node A → Node B → Node A\"\n}\n```\n\nA cycle is a **warning**, not a hard error (n8n-mcp ≥ 2.63.0) — runtime-controlled loops (error-retry, data-driven pagination, a router feeding back) execute to completion and are legitimate. **Fix** only if the loop is unintentional: ensure the cycle has a real exit (a conditional node, an error output, or a bounded counter) so it can't spin forever.\n\n#### 3. Multiple Start Nodes\n```json\n{\n  \"warning\": \"Multiple trigger nodes found - only one will execute\"\n}\n```\n\n**Fix**: Remove extra triggers or split into separate workflows\n\n#### 4. Disconnected Nodes\n```json\n{\n  \"warning\": \"Node 'Transform' is not connected to workflow flow\"\n}\n```\n\n**Fix**: Connect node or remove if unused\n\n---\n\n## Recovery Strategies\n\n### Strategy 1: Start Fresh\n**When**: Configuration is severely broken\n\n**Steps**:\n1. Note required fields from `get_node`\n2. Create minimal valid configuration\n3. Add features incrementally\n4. Validate after each addition\n\n### Strategy 2: Binary Search\n**When**: Workflow validates but executes incorrectly\n\n**Steps**:\n1. Remove half the nodes\n2. Validate and test\n3. If works: problem is in removed nodes\n4. If fails: problem is in remaining nodes\n5. Repeat until problem isolated\n\n### Strategy 3: Clean Stale Connections\n**When**: \"Node not found\" errors\n\n**Steps**:\n```javascript\nn8n_update_partial_workflow({\n  id: \"workflow-id\",\n  operations: [{\n    type: \"cleanStaleConnections\"\n  }]\n})\n```\n\n### Strategy 4: Use Auto-fix\n**When**: Validation errors that can be automatically resolved\n\n**Steps**:\n```javascript\n\u002F\u002F Preview fixes (default - doesn't apply)\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  applyFixes: false,\n  confidenceThreshold: \"medium\"  \u002F\u002F high, medium, low\n})\n\n\u002F\u002F Review fixes, then apply\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  applyFixes: true\n})\n```\n\n---\n\n## Auto-Fix Capabilities\n\nThe `n8n_autofix_workflow` tool can fix these issue types:\n\n1. **expression-format** - Missing `=` prefix in expressions (e.g., `{{ $json.field }}` → `={{ $json.field }}`)\n2. **typeversion-correction** - Downgrades nodes with unsupported typeVersions\n3. **error-output-config** - Removes conflicting onError settings\n4. **node-type-correction** - Fixes unknown node types using similarity matching (90%+ confidence)\n5. **webhook-missing-path** - Generates UUIDs for webhook nodes missing path configuration\n6. **typeversion-upgrade** - Smart upgrades to latest node versions with auto-migration\n7. **version-migration** - Guidance for complex breaking changes requiring manual steps\n\n**Confidence levels**: `high` (90%+, safe to auto-apply), `medium` (70-89%, review recommended), `low` (\u003C70%, manual review required)\n\n```javascript\n\u002F\u002F Preview all fixes\nn8n_autofix_workflow({id: \"workflow-id\"})\n\n\u002F\u002F Only apply high-confidence fixes\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  applyFixes: true,\n  confidenceThreshold: \"high\"\n})\n\n\u002F\u002F Target specific fix types\nn8n_autofix_workflow({\n  id: \"workflow-id\",\n  fixTypes: [\"expression-format\", \"typeversion-upgrade\"],\n  applyFixes: true\n})\n```\n\n**Post-update guidance**: For version upgrades, check the `postUpdateGuidance` field in the response for step-by-step migration instructions.\n\n---\n\n## Best Practices\n\n### ✅ Do\n\n- Validate after every significant change\n- Read error messages completely\n- Fix errors iteratively (one at a time)\n- Use `runtime` profile for pre-deployment\n- Check `valid` field before assuming success\n- Trust auto-sanitization for operator issues\n- Use `get_node` when unclear about requirements\n- Document false positives you accept\n\n### ❌ Don't\n\n- Skip validation before activation\n- Try to fix all errors at once\n- Ignore error messages\n- Use `strict` profile during development (too noisy)\n- Assume validation passed (always check result)\n- Manually fix auto-sanitization issues\n- Deploy with unresolved errors\n- Ignore all warnings (some are important!)\n\n---\n\n## Running the workflow after it validates\n\n`validate_workflow` checks structure, parameters and expressions — it never runs anything. A workflow that validates cleanly can still fail on real data, so run it once before calling it done.\n\n**With a webhook, form or chat trigger:** `n8n_test_workflow({workflowId})` — the default `method: \"auto\"` detects the trigger and fires it over HTTP (the workflow must be active).\n\n**Without such a trigger** (Manual Trigger, Schedule, sub-workflow) there is no HTTP entry point. Use the pin-data path, which runs through n8n's own MCP server (`N8N_MCP_ACCESS_TOKEN`, n8n 2.34+):\n\n1. `n8n_test_workflow({workflowId, method: \"prepare\"})` — lists the nodes that need pinned data.\n2. Build one sample item per listed node, keyed by node **name**, each item wrapped in `{json: {...}}`:\n   ```json\n   {\"When clicking 'Test workflow'\": [{\"json\": {\"orderId\": \"1234\", \"email\": \"a@b.com\"}}]}\n   ```\n   A flat object instead of an array of `{json}` items is the usual mistake here.\n3. `n8n_test_workflow({workflowId, method: \"pinned\", pinData})` — runs it with that data and waits for the result.\n\n**For a quick manual run without pinned data**, `method: \"direct\"` starts a manual execution and returns as soon as it has started; poll `n8n_executions({action: \"get\", id: executionId, mode: \"error\"})` for the outcome. Nothing is pinned on a `direct` run, so every node executes and any external call it makes is real — and even `pinned` only pins trigger, credentialed and HTTP Request nodes. `executionMode: \"production\"` changes the execution context, not whether there are side effects; leave it at the default unless the user asked for a production run.\n\n**`method: \"auto\"` never runs a workflow through n8n's MCP server.** On a workflow with no external trigger it reports that fact and names `prepare`\u002F`pinned`\u002F`direct`; the routed methods only run when you ask for them by name.\n\n**Consent before the first routed run.** n8n refuses these calls for a workflow whose \"Available in MCP\" setting is off, which comes back as `WORKFLOW_NOT_EXPOSED`. Re-running with `exposeToMcp: true` turns that setting on and retries once. It is a visible, persistent setting on the workflow (and enabling it is a workflow update, so a concurrent UI edit can be overwritten) — **ask the user before passing it**. The consent flow only ever enables the setting; nothing disables it implicitly. Turning it off again is a deliberate act: `n8n_update_partial_workflow({id: workflowId, operations: [{type: \"updateSettings\", settings: {availableInMCP: false}}]})`, or the toggle in the n8n UI.\n\n**Reading the result:** a run that started and then failed comes back as `EXECUTION_FAILED` with the `executionId` — inspect it with `n8n_executions({action: \"get\", id, mode: \"error\"})` and fix from the node that threw, then validate and run again.\n\n---\n\n## Reviewing an existing workflow\n\nValidating as you build (the loop above) is for catching schema and shape errors in your own in-progress work. **Reviewing an existing workflow** — yours or one you've been handed — is a different job: the workflow already passes `validate_workflow` clean, and you're hunting for the issues validation doesn't see (silent connection bugs, injection-prone queries, dropped-item Switches, Set\u002FCode antipatterns, missing error paths). For that, pull the workflow with `n8n_get_workflow` and walk **[REVIEW_CHECKLIST.md](REVIEW_CHECKLIST.md)** — a severity-tiered audit (MUST FIX \u002F SHOULD FIX \u002F NICE TO HAVE) where every item points to the canonical skill for the fix. Run `n8n_audit_instance` alongside it to surface hardcoded secrets and unauthenticated webhooks across the whole instance.\n\n---\n\n## Detailed Guides\n\nFor comprehensive error catalogs, false positives, and workflow review:\n\n- **[ERROR_CATALOG.md](ERROR_CATALOG.md)** - Complete list of error types with examples\n- **[FALSE_POSITIVES.md](FALSE_POSITIVES.md)** - When warnings are acceptable\n- **[REVIEW_CHECKLIST.md](REVIEW_CHECKLIST.md)** - Severity-tiered audit for reviewing an existing workflow\n\n---\n\n## Summary\n\n**Key Points**:\n1. **Validation is iterative** (avg 2-3 cycles, 23s + 58s)\n2. **Errors must be fixed**, warnings are optional\n3. **Auto-sanitization** normalizes operator structures on save; validation no longer errors on the raw shape\n4. **Use runtime profile** by default; step up to `ai-friendly`\u002F`strict` for best-practice advisories\n5. **Classic false positives are fixed** (≥ 2.63.0) — remaining warnings are advisories or security\u002Fdeprecation notices, not validator mistakes\n6. **Read error messages** - they contain fix guidance\n\n**Validation Process**:\n1. Validate → Read errors → Fix → Validate again\n2. Repeat until valid (usually 2-3 iterations)\n3. Review warnings and decide if acceptable\n4. Deploy with confidence\n\n**Related Skills & Tools**:\n- n8n MCP Tools Expert - Use validation tools correctly\n- n8n Expression Syntax - Fix expression errors\n- n8n Node Configuration - Understand required fields\n- `n8n_audit_instance` - Proactive security validation (hardcoded secrets, unauthenticated webhooks, missing error handling, data retention)\n",{"name":5,"description":7},"main","98985c16dc8df1bbd7dcd1bc698cd997a093e751","N8N agent\u002Fplaybooks\u002Fn8n-validation-expert","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os","https:\u002F\u002Fgithub.com\u002F26BB","🚀 Aspiring Product Manager | CS @ GS Moze\r\n\r\n🛠️ Building 0→1 products across SaaS & clean-tech\r\n\r\n📊 Turning user feedback & data into shipped features","2026-11-05 13:34:38",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add 26BB\u002Fagent-fleet-os --skill n8n-validation-expert",{"tool":37,"command":38},"bunx","bunx skills add 26BB\u002Fagent-fleet-os --skill n8n-validation-expert",{"tool":40,"command":41},"pnpm","pnpm dlx skills add 26BB\u002Fagent-fleet-os --skill n8n-validation-expert","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002FN8N agent\u002Fplaybooks\u002Fn8n-validation-expert\u002FSKILL.md",true,[45,51,55,59],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"ERROR_CATALOG.md","REFERENCE","参考文档",23796,"https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002FN8N agent\u002Fplaybooks\u002Fn8n-validation-expert\u002FERROR_CATALOG.md",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"FALSE_POSITIVES.md",18180,"https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002FN8N agent\u002Fplaybooks\u002Fn8n-validation-expert\u002FFALSE_POSITIVES.md",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"README.md",8474,"https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002FN8N agent\u002Fplaybooks\u002Fn8n-validation-expert\u002FREADME.md",{"path":60,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":61,"sourceHtmlUrl":62},"REVIEW_CHECKLIST.md",15253,"https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002FN8N agent\u002Fplaybooks\u002Fn8n-validation-expert\u002FREVIEW_CHECKLIST.md",[64,71,78,85,92,99],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7F51NQSCRM2PWFMATCV9MEG","taste-design","Semantic Design System Skill for Google Stitch. Generates agent-friendly DESIGN.md files that enforce premium, anti-generic UI standards — strict typography, calibrated color, asymmetric layouts, perpetual micro-motion, and hardware-accelerated performance.","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002Fstitch agent\u002Fplugins\u002Fstitch-utilities\u002Fplaybooks\u002Ftaste-design\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7F51JGTS9VJ3S9T6VMQWZ08","stitch-loop","Teaches agents to iteratively build websites using Stitch with an autonomous baton-passing loop pattern","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002Fstitch agent\u002Fplugins\u002Fstitch-utilities\u002Fplaybooks\u002Fstitch-loop\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M7F51EZR2XYDWMA3XKWH1HX2","site-md","Analyze a project description or site requirements and synthesize a project constitution into SITE.md for the Stitch Build Loop","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002Fstitch agent\u002Fplugins\u002Fstitch-utilities\u002Fplaybooks\u002Fsite-md\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7F51BNCMP0GQ7SGP39SZYDP","enhance-prompt","Transforms vague UI ideas into polished, Stitch-optimized prompts. Enhances specificity, adds UI\u002FUX keywords, injects design system context, and structures output for better generation results.","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002Fstitch agent\u002Fplugins\u002Fstitch-utilities\u002Fplaybooks\u002Fenhance-prompt\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":19},"01M7F518SX707EB3X42B2V7HRA","design-md","Analyze Stitch projects and synthesize a semantic design system into DESIGN.md files","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002Fstitch agent\u002Fplugins\u002Fstitch-utilities\u002Fplaybooks\u002Fdesign-md\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":19},"01M7F5148PVYTA76NR3X0YK8SK","stitch::upload-to-stitch","Upload local assets (images, mockups, extracted HTML, design markdown) to a Stitch project. ALWAYS use this skill when you need to upload visual assets, HTML pages, or design docs to Stitch, particularly when direct MCP tool calls fail or truncate due to base64 token limits.","https:\u002F\u002Fgithub.com\u002F26BB\u002Fagent-fleet-os\u002Fblob\u002Fmain\u002Fstitch agent\u002Fplugins\u002Fstitch-design\u002Fplaybooks\u002Fupload-to-stitch\u002FSKILL.md",[],[]]