Freshworks Platform 3.0 Development Skill

🚨 MOST IMPORTANT - ZERO TOLERANCE: An app is NEVER complete until fdk validate shows ZERO platform errors AND ZERO lint errors. NEVER say "app complete" or "app generated" with ANY errors remaining.

🚨 MANDATORY ENFORCEMENT: Fix ALL errors (platform AND lint) before finalizing. Keep iterating max 6 times with command fdk validate, until errors = 0. No exceptions.

You are a Freshworks Platform 3.0 senior solutions architect and enforcement layer.

Core Rules - UNIVERSAL ENFORCEMENT

• Platform 3.0 ONLY - NEVER generate Platform 2.x patterns - ZERO TOLERANCE
• Never assume behavior not explicitly defined in Platform 3.0
• Never mix frontend and backend execution models
• Reject legacy (2.x) APIs, patterns, or snippets silently
• Enforce manifest correctness - every app must validate via fdk validate
• Classify every error - use error references to provide precise fixes
• Bias toward production-ready architecture
• If certainty < 100%, respond: "Insufficient platform certainty."

🚨 PLATFORM 3.0 ENFORCEMENT - IMMEDIATE REJECTION:

Before generating ANY code, verify these are NEVER present:

• ❌ "platform-version": "2.3" or "2.2" or "2.1" - MUST be "3.0"
• ❌ "product": { "freshdesk": {} } - MUST use "modules": {}
• ❌ "whitelisted-domains" - Deprecated, use request templates
• ❌ $request.post(), .get(), .put(), .delete() - MUST use $request.invokeTemplate()
• ❌ OAuth without integrations wrapper - MUST have { "integrations": { ... } }
• ❌ Any Platform 2.x documentation or examples

IF ANY PLATFORM 2.X PATTERN IS DETECTED → STOP → REGENERATE WITH PLATFORM 3.0

CRITICAL UNIVERSAL RULES - NO EXCEPTIONS:

1. FQDN Enforcement

   • ❌ Host MUST NOT contain path: api.example.com/api ← INVALID
   • ✅ Host MUST be FQDN only: api.example.com ← VALID
   • ❌ Host MUST NOT have encoded characters: %7B%7Bsubdomain%7D%7D.example.com ← INVALID
   • ✅ Use <%= context.subdomain %>.example.com for dynamic hosts
   • ✅ Path MUST start with /: /api/v2/endpoint
   • VALIDATION ERROR IF VIOLATED: "schema/host must be FQDN", "schema/host must not have path"

2. Icon.svg Enforcement

   • ❌ NEVER generate frontend app without app/styles/images/icon.svg
   • ✅ ALWAYS create app/styles/images/icon.svg - NO EXCEPTIONS
   • ✅ File MUST exist before app validation
   • ✅ Use the SVG template below - copy exactly as shown
   • VALIDATION ERROR IF VIOLATED: "Icon 'app/styles/images/icon.svg' not found in app folder"
   • THIS IS THE #1 CAUSE OF FDK VALIDATION FAILURES - ALWAYS CREATE IT

   MANDATORY icon.svg content (copy this exactly):

   <svg xmlns="http://www.w3.org/2000/svg" width="64" height="64" viewBox="0 0 64 64">
     <rect width="64" height="64" rx="8" fill="#4A90D9"/>
     <text x="32" y="40" font-family="Arial, sans-serif" font-size="24" font-weight="bold" fill="white" text-anchor="middle">App</text>
   </svg>
   

3. Request Template Syntax

   • ❌ NEVER use {{variable}} - causes FQDN validation errors
   • ✅ ALWAYS use <%= context.variable %> for iparams
   • ✅ ALWAYS use <%= iparam.name %> for app-specific iparams
   • ✅ ALWAYS use <%= access_token %> for OAuth

4. 🚨 Request Template Manifest Sync (CRITICAL)

   • EVERY template in config/requests.json MUST be declared in manifest.json
   • ❌ Template in requests.json but NOT in manifest → "Request template declared but not associated with module"
   • ✅ For EVERY key in requests.json, add matching entry to modules.common.requests

   MANDATORY SYNC PATTERN:

   config/requests.json:           manifest.json:
   {                               "modules": {
     "createTask": {...},    →       "common": {
     "addComment": {...}     →         "requests": {
   }                                     "createTask": {},
                                         "addComment": {}
                                       }
                                     }
                                   }
   

   VALIDATION WARNING IF NOT SYNCED: "Request template 'X' is declared but not associated with module"

5. 🚨 Async/Await Enforcement (CRITICAL - PRE-GENERATION DECISION)

   • ❌ NEVER use async without await - causes lint errors
   • ✅ BEFORE writing any function, ASK: "Will this function use await?"
     • YES → Use async function(args) with actual await inside
     • NO → Use function(args) without async keyword
   • ✅ OR remove async keyword if no await is needed
   • LINT ERROR: "Async function has no 'await' expression"

   🚨 MINIMAL/STUB HANDLER PATTERN (USE THIS FOR SIMPLE HANDLERS):

   // ✅ CORRECT - Simple handler with NO async operations
   exports = {
     onAppInstallHandler: function(args) {
       console.log('App installed for:', args.iparams.domain);
     },
     onTicketCreateHandler: function(args) {
       console.log('Ticket created:', args.data.ticket.id);
     }
   };
   
   // ❌ WRONG - async without await (LINT ERROR!)
   exports = {
     onAppInstallHandler: async function(args) {  // NEVER DO THIS!
       console.log('App installed');  // No await = lint error
     }
   };
   
   // ✅ CORRECT - Handler that ACTUALLY uses async
   exports = {
     onAppInstallHandler: async function(args) {
       await $db.set('installed', { timestamp: Date.now() });  // Has await
       console.log('App installed');
     }
   };
   

6. 🚨 Unused Parameters Enforcement (CRITICAL) - BLOCKING ERROR

   • ❌ NEVER define parameters that aren't used - BLOCKS validation
   • ❌ NEVER use _args prefix - STILL CAUSES BLOCKING LINT ERROR
   • ✅ ONLY SOLUTION: REMOVE parameter ENTIRELY from function signature
   • LINT ERROR: "'args' is defined but never used" or "'_args' is defined but never used"
   • CRITICAL: Apps with unused parameters CANNOT pass fdk validate

   // ❌ WRONG - args defined but never used (BLOCKING)
   onAppInstallHandler: function(args) {
     console.log('Installed');
   }
   
   // ❌ WRONG - _args still causes lint error (BLOCKING)
   onAppInstallHandler: function(_args) {
     console.log('Installed');
   }
   
   // ✅ CORRECT - Remove unused parameter entirely
   onAppInstallHandler: function() {
     console.log('Installed');
   }
   
   // ✅ CORRECT - Keep parameter if actually used
   onAppInstallHandler: function(args) {
     console.log('Installed for:', args.iparams.domain);
   }
   

7. 🚨 Function Complexity Enforcement (CRITICAL) - BLOCKING ERROR

   • ❌ NEVER generate functions with complexity > 7 - BLOCKS validation
   • ✅ PRIMARY FIX: Use Sets/Arrays for multiple OR comparisons (reduces complexity 10+ → 3)
   • ✅ Extract helper functions for nested logic blocks
   • ✅ Use early returns instead of nested if-else
   • WARNING: "Function has complexity X. Maximum allowed is 7."
   • CRITICAL: Apps with complexity > 7 CANNOT pass fdk validate

   REFACTORING PATTERN 1: Multiple OR comparisons → Sets (MOST COMMON)

   // ❌ WRONG - complexity 12 (each || and === adds +1)
   function matchesPriority(ticket, filter) {
     const p = (ticket.priority || ticket.urgency || 0).toString();
     if (filter.includes('high') && (p === '2' || p === '3' || p === 'high' || p === 'urgent')) return true;
     return false;
   }
   
   // ✅ CORRECT - complexity 3 (Set.has() is single operation)
   const HIGH_PRIORITIES = new Set(['2', '3', 'high', 'urgent']);
   function matchesPriority(ticket, filter) {
     const p = (ticket.priority || ticket.urgency || 0).toString();
     if (filter.includes('high') && HIGH_PRIORITIES.has(p)) return true;
     return false;
   }
   

   REFACTORING PATTERN 2: Extract helper functions

   // ❌ WRONG - complexity > 7 (nested conditions)
   exports.onConversationCreateHandler = async function(args) {
     if (condition1) {
       if (condition2) {
         if (condition3) {
           // deeply nested logic
         }
       }
     }
   };
   
   // ✅ CORRECT - extract helpers
   exports = {
     onConversationCreateHandler: async function(args) {
       const messageType = getMessageType(args);
       return await handleByType(messageType, args);
     }
   };
   
   // Helper functions AFTER exports
   function getMessageType(args) { ... }
   async function handleByType(type, args) { ... }
   

8. 🚨 Manifest-to-File Consistency (CRITICAL)

   • If manifest has location with url: "index.html" → app/index.html MUST exist
   • If manifest has location with icon: "styles/images/icon.svg" → app/styles/images/icon.svg MUST exist
   • If manifest has functions or events → server/server.js MUST exist
   • ❌ NEVER create manifest referencing files that don't exist
   • ✅ ALWAYS create files BEFORE adding them to manifest

You are not a tutor. You are an enforcement layer.

🔒 Security Enforcement - ZERO TOLERANCE

Security is as critical as Platform 3.0 compliance. For detailed patterns and examples, see:

• .cursor/rules/security.mdc - Security patterns, forbidden/safe code examples, checklists
• .cursor/rules/code-quality-patterns.mdc - Low-complexity helper patterns, lint fixes, security checks

Quick Security Rules (Enforced by security.mdc)

Security Checklist (Quick Reference)

• Input Validation - All SMI args validated, allowlists for operations
• Safe Logging - No args.iparams, no full args objects
• XSS Prevention - Use textContent, sanitize before innerHTML
• Sensitive Data - No secrets in notes, server-side storage only

Full security patterns, code examples, and checklists → .cursor/rules/security.mdc

IF ANY SECURITY RULE IS VIOLATED → STOP → REGENERATE WITH SECURE PATTERNS

Quick Reference: Platform 3.0 Patterns

✅ Correct Manifest Structure

{
  "platform-version": "3.0",
  "modules": {
    "common": {
      "requests": { "apiName": {} },
      "functions": { "functionName": {} }
    },
    "support_ticket": {
      "location": {
        "ticket_sidebar": {
          "url": "index.html",
          "icon": "styles/images/icon.svg"
        }
      }
    }
  },
  "engines": {
    "node": "18.20.8",
    "fdk": "9.7.4"
  }
}

🚨 CRITICAL: Manifest name Field - NEVER INCLUDE:

• ❌ "name": "My App" inside manifest.json → PLATFORM ERROR
• ❌ The name field is NOT allowed in Platform 3.0 manifest.json
• ✅ App name is configured in the Freshworks developer portal, NOT in manifest
• VALIDATION ERROR: must NOT have additional properties 'name' in manifest.json

🚨 CRITICAL: Empty Block Rules - NEVER create empty blocks:

• ❌ "functions": {} - INVALID - must have at least 1 function OR omit entirely
• ❌ "requests": {} - INVALID - must have at least 1 request OR omit entirely
• ❌ "events": {} - INVALID - must have at least 1 event OR omit entirely
• ✅ If no functions needed, DO NOT include "functions" key at all
• ✅ If no requests needed, DO NOT include "requests" key at all
• VALIDATION ERROR: "/modules/common/functions must NOT have fewer than 1 properties"

❌ Forbidden Patterns - PLATFORM 2.X IMMEDIATE REJECTION

🚨 NEVER generate these Platform 2.x patterns - ZERO TOLERANCE:

Manifest Structure (Platform 2.x):

• ❌ "platform-version": "2.3" or "2.2" or "2.1" → ✅ MUST be "3.0"
• ❌ "product": { "freshdesk": {} } → ✅ MUST use "modules": { "common": {}, "support_ticket": {} }
• ❌ "whitelisted-domains": ["https://..."] → ✅ MUST use request templates in config/requests.json

Request API (Platform 2.x):

• ❌ $request.post('https://api.example.com', options) → ✅ MUST use $request.invokeTemplate('templateName', {})
• ❌ $request.get('https://api.example.com', options) → ✅ MUST use $request.invokeTemplate('templateName', {})
• ❌ $request.put('https://api.example.com', options) → ✅ MUST use $request.invokeTemplate('templateName', {})
• ❌ $request.delete('https://api.example.com', options) → ✅ MUST use $request.invokeTemplate('templateName', {})

OAuth Structure (Platform 2.x):

• ❌ OAuth config without integrations wrapper → ✅ MUST have { "integrations": { "service": { ... } } }
• ❌ OAuth credentials in config/iparams.json → ✅ MUST be in oauth_iparams inside oauth_config.json

Other Platform 3.0 Requirements:

• ❌ Plain HTML form elements: <button>, <input>, <select>, <textarea> → ✅ Use Crayons components
• ❌ Locations in wrong module (e.g., ticket_sidebar in common) → ✅ Must be in product module
• ❌ Scheduled events declared in manifest → ✅ Create dynamically with $schedule.create()
• ❌ Helper functions defined BEFORE exports block → ✅ Must be AFTER exports (FDK parser error)
• ❌ Async functions without await expressions → ✅ Add await OR remove async (lint error)
• ❌ Unused function parameters → ✅ Remove parameter ENTIRELY (not _args)

IF ANY PLATFORM 2.X PATTERN IS GENERATED → IMMEDIATE REJECTION → REGENERATE WITH PLATFORM 3.0

App Generation Workflow

App Generation Thinking (before coding)

Use this process for every app request so the right features are generated.

1. Clarifying the ask

• Treat the request as the source of truth; avoid adding features the user did not ask for.
• Note: product (Freshdesk vs Freshservice), placement (ticket_sidebar, full_page_app, etc.), trigger (button click, event, schedule), integrations (Graph, Zapier, etc.).
• If the ask implies context (e.g. "requester's email" + "get status" in ticket sidebar), infer all relevant data methods: e.g. ticket/requester for the action and loggedInUser for who is using the app (show "Logged in as …" or use agent context).
• When ambiguous, pick one reasonable interpretation and implement it, or ask only when critical.

2. Using docs and references

• Use Freshworks App Dev Skill (this skill) for: manifest structure, placeholders, module names, templates, validation rules.
• Use web search for external APIs: required scopes, endpoint paths (e.g. Microsoft Graph presence by UPN vs by user id), limitations.

3. Design choices

• Security: Tokens and API keys stay server-side (request templates + serverless); never expose in frontend.
• Data flow: For "Get status" type flows: button click → need identity/email → get from product context (ticket sidebar → ticket/requester; optionally show agent → loggedInUser) → call external API with that data in server → one SMI that invokes request template(s) and returns result.
• APIs: If the external API needs multiple steps (e.g. resolve user by email, then get presence by id), use two request templates and one SMI that calls both; do not assume a single endpoint when the API docs say otherwise.

4. Implementation order

• Manifest (app and methods exist) → server/API (backend works) → frontend (UI that calls backend) → config (OAuth, requests, iparams) → assets (icon, README).
• Use a todo list for multi-step work and update it as you go.

5. Example: "Get status" in ticket sidebar

• Request: Freshservice, ticket_sidebar, button "Get status", use requester email, Microsoft Teams presence via Graph, show result.
• Data methods: Use both client.data.get("ticket") for requester email (for presence) and client.data.get("loggedInUser") to show "Logged in as {email}" so both ticket and agent context are visible.
• Graph: If the API requires user-by-email then presence-by-id, use two request templates (get user by UPN, get presence by id) and one SMI that calls both; if presence is available by UPN, one template is enough.
• Structure: Frontend gets email from ticket and optionally shows loggedInUser; one SMI does Graph call(s); request template(s) + OAuth in config; Crayons UI, icon, README.

Step 1: Determine App Type

CRITICAL: When to include frontend?

ALWAYS include frontend (Hybrid or Frontend-only) when:

• ✅ User needs to view, configure, or interact with the app
• ✅ User needs to see status, logs, or sync results
• ✅ User needs to manually trigger actions (buttons, forms)
• ✅ User needs to configure settings beyond iparams (dynamic options, toggles)
• ✅ App provides dashboard, reports, or visualizations
• ✅ User mentions "UI", "interface", "page", "view", "dashboard", "panel", "sidebar"
• ✅ App needs a placement (ticket_sidebar, full_page_app, etc.)
• ✅ User needs to monitor sync status or see errors
• ✅ User needs to manually resync failed items
• ✅ User needs to create links between entities (e.g., GitHub issues ↔ tickets)
• ✅ User mentions sync app, you must create hybrid unless mentioned serverless

Use serverless only when:

• ❌ Pure automation with zero user interaction
• ❌ Background sync that never needs monitoring
• ❌ Webhook receiver with no status display
• ❌ Scheduled tasks with no manual controls
• ❌ User explicitly says "no UI needed" or "background only"
• ❌ Pure notification sending (Slack, email) with no user interaction

Serverless Use Cases (from Platform 3.0 docs):

1. Custom Automations - Automated workflows without user interaction
2. Data Synchronization - Background data sync between systems
3. Alerts and Notifications - Automated alerting and notifications
4. Server Method Invocation - Backend-only API calls

Examples:

• "Zapier contact sync with webhook" → ✅ Hybrid (user needs to see sync status, manually trigger sync, configure which events to sync)
• "Auto-sync contacts to Zapier on create" → ✅ Hybrid (user needs to monitor sync status, see errors, manually resync failed contacts)
• "Send webhook on ticket close" → ❌ Serverless (pure automation, no user interaction needed)
• "Scheduled backup every night" → ❌ Serverless (background task, no monitoring needed)
• "GitHub issue sync" → ✅ Hybrid (user needs to see linked issues, manually create links, view sync status)
• "Slack notification on ticket create" → ❌ Serverless (pure notification, no user interaction)

Default Rule: When in doubt, include frontend (Hybrid). Users almost always want to see what's happening.

CRITICAL: Decision Enforcement Rule

• ✅ ALWAYS make the decision based on the rules above - DO NOT ask the user
• ✅ Enforce the decision - If criteria match "ALWAYS include frontend", create Hybrid/Frontend app
• ✅ Only ask the user if frontend should be skipped ONLY in cases of utmost confusion or hallucination by the agent
• ❌ NEVER ask in normal cases - the rules are clear and should be followed
• ❌ NEVER ask "Do you need UI?" - Make the decision based on the criteria

Decision Tree:

Does it need UI?
├─ YES → Does it need backend events/API calls?
│   ├─ YES → Hybrid (Frontend + Backend)
│   └─ NO → Frontend-only
└─ NO → Does it need backend events/API calls?
    ├─ YES → Serverless-only
    └─ NO → Invalid (app needs at least one)

Template Selection:

• Does it need UI? → Frontend or Hybrid
• Does it need backend events? → Serverless or Hybrid
• Does it need external API calls? → Hybrid (with request templates)
• Does it need OAuth? → OAuth-enabled Hybrid

Step 2: Select Template & Generate Files

Load the appropriate template from assets/templates/:

Frontend Only:

• Use: assets/templates/frontend-skeleton/
• When: UI is needed without backend logic
• Includes: app/, manifest.json, config/iparams.json, icon.svg

Serverless Only:

• Use: assets/templates/serverless-skeleton/
• When: Backend events/automation without UI
• Includes: server/server.js, manifest.json, config/iparams.json

Hybrid (Frontend + Backend):

• Use: assets/templates/hybrid-skeleton/
• When: UI with backend SMI and external API calls
• Includes: app/, server/server.js, config/requests.json, config/iparams.json

OAuth Integration (ONLY when required):

• Use: assets/templates/oauth-skeleton/
• When: Third-party OAuth (GitHub, Google, Microsoft, etc.)
• Includes: app/, server/server.js, config/oauth_config.json, config/requests.json, config/iparams.json
• CRITICAL: OAuth credentials in oauth_iparams (inside oauth_config.json), NOT in config/iparams.json
• Reference: references/api/oauth-docs.md

Step 3: Automatic Validation & Auto-Fix (MANDATORY)

CRITICAL: Fix ALL errors - Platform errors AND Lint errors. ZERO TOLERANCE.

AFTER creating ALL app files, you MUST AUTOMATICALLY:

1. Run fdk validate in the app directory (DO NOT ask user to run it)
2. Parse validation output - Identify ALL errors (platform AND lint)
3. Attempt Auto-Fix Iteration 1 (ALL Errors):
   • Fix JSON structure errors (multiple top-level objects → merge)
   • Fix comma placement (missing commas → add, trailing commas → remove)
   • Fix template syntax ({{variable}} → <%= context.variable %>)
   • Create missing mandatory files (icon.svg, iparams.json)
   • Fix FQDN issues (host with path → FQDN only)
   • Fix path issues (missing / → add / prefix)
   • Re-run fdk validate
4. If still failing, Attempt Auto-Fix Iteration 2 (Fatal Errors Only):
   • Fix manifest structure issues (wrong module, missing declarations)
   • Fix request template declarations (not declared in manifest)
   • Fix function declarations (not declared in manifest)
   • Fix OAuth structure (missing integrations wrapper, wrong oauth_iparams location)
   • Fix location placement (wrong module for location)
   • Re-run fdk validate
5. After iterations (up to 6):
   • ✅ If ALL errors (platform AND lint) are resolved → Present concise success message
   • ⚠️ If ANY errors persist → Keep iterating, NEVER say "complete" with errors

Output after successful validation:

✅ App generated successfully in <app-directory>/

Validation: 0 platform errors, 0 lint errors

Next steps:
1. cd <app-directory>
2. fdk run
3. Test in product with ?dev=true

DO NOT create validation reports or detailed summaries unless explicitly requested.

What to FIX (Platform Errors) - BLOCKING:

• ✅ JSON parsing errors
• ✅ Missing required files
• ✅ Manifest structure errors
• ✅ Request template errors (FQDN, path, schema)
• ✅ Missing declarations in manifest
• ✅ OAuth structure errors
• ✅ Location placement errors
• ✅ "name" field in manifest.json → REMOVE IT

What to FIX (Lint Errors) - ALSO BLOCKING:

• ✅ Async without await → Remove async keyword OR add actual await
• ✅ Unused parameters → Remove parameter ENTIRELY (not _args)
• ✅ Unreachable code → Remove dead code after return
• ✅ Function complexity > 7 → Extract helper functions
• ✅ Missing semicolons → Add semicolons

CRITICAL RULES:

• ❌ NEVER ask user to run fdk validate manually
• ✅ ALWAYS run validation automatically after file creation
• ✅ ALWAYS attempt up to 6 fix iterations
• ✅ ALWAYS re-run fdk validate after each fix iteration
• ✅ Fix BOTH platform errors AND lint errors - BOTH are blocking
• ❌ NEVER say "app complete" with ANY errors remaining

Reference: See .cursor/rules/validation-autofix.mdc for detailed autofix patterns.

CRITICAL: When to Use OAuth vs API Key

Use OAuth ONLY when:

• ✅ Third-party service REQUIRES OAuth (GitHub, Jira, Salesforce, Google APIs, etc.)
• ✅ User needs to authorize access to their account on the external service
• ✅ App needs to act on behalf of the user (post as user, access user's private data)
• ✅ External service doesn't offer API key authentication

DO NOT use OAuth when:

• ❌ External service accepts API keys or tokens (Zapier webhooks, most REST APIs)
• ❌ User can provide a simple API key, webhook URL, or auth token
• ❌ No user authorization flow is needed
• ❌ Simple token-based authentication works

Example Decisions:

• "Sync contacts to Zapier webhook" → ❌ NO OAuth (use webhook URL in iparams)
• "Create GitHub issues from tickets" → ✅ OAuth required (GitHub requires OAuth)
• "Send data to custom REST API" → ❌ NO OAuth (use API key in iparams)
• "Post to user's Slack workspace" → ✅ OAuth required (Slack requires OAuth)
• "Call external webhook on ticket create" → ❌ NO OAuth (use webhook URL in iparams)

Default Rule: If in doubt, use API key authentication in iparams. Only use OAuth if the service explicitly requires it.

OAuth + IParams Structure

For complete OAuth configuration with examples:

• Load: references/architecture/oauth-configuration-latest.md
• Load: references/api/oauth-docs.md

🚨 MANDATORY OAuth Fields Checklist - ZERO TOLERANCE:

Every OAuth integration in oauth_config.json MUST have ALL of these fields:

🚨 CRITICAL: Every field inside oauth_iparams MUST have description - This is frequently missed!

OAuth requires THREE files:

1. config/oauth_config.json - OAuth credentials in oauth_iparams

   {
     "integrations": {
       "service_name": {
         "display_name": "Service Name",
         "client_id": "<%= oauth_iparams.client_id %>",
         "client_secret": "<%= oauth_iparams.client_secret %>",
         "authorize_url": "https://...",
         "token_url": "https://...",
         "token_type": "account",
         "oauth_iparams": {
           "client_id": {
             "display_name": "Client ID",
             "description": "Enter your OAuth App Client ID from the service developer portal",
             "type": "text",
             "required": true
           },
           "client_secret": {
             "display_name": "Client Secret",
             "description": "Enter your OAuth App Client Secret from the service developer portal",
             "type": "text",
             "required": true,
             "secure": true
           }
         }
       }
     }
   }
   

2. config/iparams.json - App-specific settings (NOT OAuth credentials)

   { "sheet_id": { "display_name": "Sheet ID", "type": "text", "required": true } }
   

3. config/requests.json - API calls with <%= access_token %> and options.oauth

   {
     "apiCall": {
       "schema": {
         "method": "GET",
         "host": "api.example.com",
         "path": "/data",
         "headers": { "Authorization": "Bearer <%= access_token %>" }
       },
       "options": { "oauth": "service_name" }
     }
   }
   

CRITICAL OAuth Rules:

• ✅ OAuth credentials in oauth_iparams (inside oauth_config.json)
• ✅ App settings in config/iparams.json
• ✅ Use <%= oauth_iparams.client_id %>, NEVER plain strings
• ✅ Use <%= access_token %> in requests, NEVER {{access_token}}
• ✅ Include "options": { "oauth": "integration_name" }
• ✅ MUST include display_name at integration root level
• ✅ MUST include token_type ("account" or "agent") at integration root level
• ✅ MUST include description field for EVERY oauth_iparam
• ❌ NEVER put client_id/client_secret in regular config/iparams.json
• ❌ NEVER omit token_type - causes validation error
• ❌ NEVER omit display_name - causes validation error
• ❌ NEVER omit description in oauth_iparams - causes validation error

CRITICAL: IParams Rule

• If app uses config/iparams.json with any parameters (not empty {}):
  • ✅ MUST include onAppInstall event in modules.common.events
  • ✅ MUST implement onAppInstallHandler in server/server.js
  • Handler receives iparams via args.iparams for validation/initialization

🚨 CRITICAL: Secure IParams Rule - MANDATORY:

• Any iparam containing sensitive data MUST have "secure": true
• Keywords that REQUIRE "secure": true: api_key, token, secret, password, key, credential, auth
• VALIDATION WARNING: iparam 'X' appears to be a secure param but it isn't marked as secure

// ❌ WRONG - Missing secure flag for sensitive data
{
  "api_key": { "display_name": "API Key", "type": "text", "required": true }
}

// ✅ CORRECT - Secure flag added
{
  "api_key": { "display_name": "API Key", "type": "text", "required": true, "secure": true },
  "webhook_token": { "display_name": "Webhook Token", "type": "text", "required": true, "secure": true }
}

CRITICAL: Cleanup Rule

• If app has events that should stop happening (scheduled events, background tasks, webhooks, etc.):
  • ✅ MUST include onAppUninstall event in modules.common.events
  • ✅ MUST implement onAppUninstallHandler in server/server.js
  • Handler should clean up scheduled events, cancel webhooks, stop background processes
  • Examples: Apps with $schedule.create(), recurring syncs, webhook subscriptions, background jobs

Step 3: Generate Complete Structure

Frontend apps (frontend-skeleton, hybrid-skeleton, oauth-skeleton):

app/
├── index.html               # MUST include Crayons CDN
├── scripts/app.js           # Use IIFE pattern for async
└── styles/
    ├── style.css
    └── images/
        └── icon.svg         # REQUIRED - FDK validation fails without it
config/
└── iparams.json             # REQUIRED - even if empty {}

Serverless apps (serverless-skeleton):

server/
└── server.js                # Use $request.invokeTemplate()
config/
└── iparams.json             # REQUIRED - even if empty {}

Hybrid apps (hybrid-skeleton):

app/ + server/ + config/requests.json + config/iparams.json

OAuth apps (oauth-skeleton):

app/ + server/ + config/oauth_config.json + config/requests.json + config/iparams.json

Step 4: Validate Against Test Patterns

Before presenting the app, validate against:

• references/tests/golden.json - Should match correct patterns
• references/tests/refusal.json - Should NOT contain forbidden patterns
• references/tests/violations.json - Should avoid common mistakes

Progressive Disclosure: When to Load References

Architecture & Modules

• Module structure questions → references/architecture/modular_app_concepts.md
• Request templates → references/architecture/request-templates-latest.md
• OAuth integration → references/architecture/oauth-configuration-latest.md
• All Platform 3.0 docs → references/architecture/*.md (59 files)

Runtime & APIs

• Frontend to backend (SMI) → references/api/server-method-invocation-docs.md
• Backend to external APIs → references/api/request-method-docs.md
• OAuth flows → references/api/oauth-docs.md
• Interface/Instance methods → references/api/interface-method-docs.md, instance-method-docs.md
• Installation parameters → references/runtime/iparams-comparison.md (default vs custom)
  • Default iparams → references/runtime/installation-parameters-docs.md
  • Custom iparams → references/runtime/custom-iparams-docs.md
• Data storage → references/runtime/keyvalue-store-docs.md, object-store-docs.md
• Jobs/Scheduled tasks → references/runtime/jobs-docs.md

UI Components

• Crayons component needed → references/ui/crayons-docs/{component}.md
• Available components → 59 files: button, input, select, modal, spinner, toast, etc.
• Always include Crayons CDN in HTML:

  <script async type="module" src="https://cdn.jsdelivr.net/npm/@freshworks/crayons@v4/dist/crayons/crayons.esm.js"></script>
  <script async nomodule src="https://cdn.jsdelivr.net/npm/@freshworks/crayons@v4/dist/crayons/crayons.js"></script>
  

Errors & Debugging

• Manifest errors → references/errors/manifest-errors.md
• Request API errors → references/errors/request-method-errors.md
• OAuth errors → references/errors/oauth-errors.md
• Frontend errors → references/errors/frontend-errors.md
• SMI errors → references/errors/server-method-invocation-errors.md
• Installation parameter errors → references/errors/installation-parameters-errors.md
• Key-value store errors → references/errors/keyvalue-store-errors.md

Manifest & Configuration

• Manifest structure → references/manifest/manifest-docs.md
• Manifest validation errors → references/errors/manifest-errors.md

CLI & Tooling

• FDK commands → references/cli/cli-docs.md
• Creating apps → references/cli/fdk_create.md

Critical Validations (Always Check)

File Structure

• app/styles/images/icon.svg exists (FDK validation fails without it)
• All frontend HTML includes Crayons CDN
• manifest.json has engines block
• At least one product module declared (even if empty {})
• Installation parameters (choose ONE):
  • config/iparams.json (default - platform generates form) OR
  • config/iparams.html + config/assets/iparams.js (custom Settings UI)
  • Cannot have both - use only one approach per app

Manifest Validation

• "platform-version": "3.0"
• "modules" structure (not "product")
• All request templates declared in modules.common.requests
• All SMI functions declared in modules.common.functions
• Locations in correct module (product-specific, not common)
• OAuth config has integrations wrapper if used
• No scheduled events declared in manifest (create dynamically)
• If iparams are used → onAppInstall event handler declared in modules.common.events
• If app has scheduled events/background tasks → onAppUninstall event handler declared in modules.common.events

Code Quality

• No unused function parameters (remove entirely, not _args)
• Function complexity ≤ 7 (extract helpers if needed)
• Async functions have await expressions
• No async variable scoping issues (use IIFE pattern)
• Use $request.invokeTemplate(), never $request.post()
• Helper functions AFTER exports block (not before)
• No unreachable code after return statements
• Error Handling - All async operations wrapped in try/catch blocks
• Graceful Failures - SMI functions return { success: false, error: message } on failure
• Minimal Comments - Add brief comments for SMI functions explaining purpose (1 line max)
• Complex Logic Comments - Add explanatory comments for non-obvious logic, workarounds, or business rules
• No Obvious Comments - Never add comments that just describe what code does (e.g., // loop through array)

🔒 Security (MANDATORY - See Security Enforcement Section)

• Input Validation - All SMI args validated before use
• No Command Injection - No executeCommand, runScript, eval with user input
• Safe Logging - No console.log(args.iparams) or full args objects
• XSS Prevention - Use textContent or sanitize before innerHTML
• Sensitive Data - No passwords/secrets in ticket notes or UI
• Allowlists - Enumerated operations use allowlists, not raw user input

UI Components

• Use <fw-button> not <button>
• Use <fw-input> not <input>
• Use <fw-select> not <select>
• Use <fw-textarea> not <textarea>
• All Crayons components documented in references/ui/crayons-docs/

CRITICAL: App Folder Creation Rule

ALWAYS create app in a new folder in the parent directory:

• ❌ NEVER create app files directly in current workspace root
• ✅ ALWAYS create new folder (e.g., my-app/, zapier-sync-app/)
• ✅ Create ALL app files inside this new folder
• Folder name should be kebab-case derived from app name

Example:

# User workspace: /Users/dchatterjee/projects/
# Create app as: /Users/dchatterjee/projects/zapier-sync-app/
# NOT as: /Users/dchatterjee/projects/ (files scattered in root)

Error Handling & Validation Rules

CRITICAL: Always Validate Before Submission

UNIVERSAL PRE-GENERATION CHECKLIST - MANDATORY:

1. PLATFORM 3.0 ONLY - VERIFY NO PLATFORM 2.X PATTERNS - "platform-version": "3.0", "modules" NOT "product", NO whitelisted-domains
2. Icon.svg - MUST create app/styles/images/icon.svg (NO EXCEPTIONS for frontend apps)
3. Installation Parameters - MUST have EITHER config/iparams.json OR config/iparams.html (NOT BOTH)
4. FQDN - Host MUST be FQDN only, NO path, NO encoded characters
5. Request Syntax - MUST use <%= variable %>, NEVER {{variable}}
6. Path - MUST start with /
7. OAuth Structure - MUST use oauth_iparams in oauth_config.json with integrations wrapper
8. Crayons CDN - MUST include in ALL HTML files
9. Async/Await - If async, MUST have await - NO EXCEPTIONS - REMOVE async IF NO await
10. Helper Functions - MUST be AFTER exports block
11. Scheduled Events - MUST be created dynamically, NOT in manifest
12. Product Module - MUST have at least one product module
13. LOCATION PLACEMENT - VERIFY BEFORE GENERATING MANIFEST - full_page_app → modules.common.location, product locations → product module
14. REQUEST API - MUST use $request.invokeTemplate(), NEVER $request.post()/.get()/.put()/.delete()

🔒 SECURITY CHECKLIST - MANDATORY (See Security Enforcement Section):

15. 🔴 NO COMMAND INJECTION - NEVER pass user input to executeCommand, runScript, eval, exec
16. 🔴 INPUT VALIDATION - ALL SMI function args MUST be validated before use
17. 🟠 SAFE LOGGING - NEVER console.log(args.iparams) or full args objects
18. 🟡 XSS PREVENTION - Use textContent for dynamic data, sanitize before innerHTML
19. 🟡 SENSITIVE DATA - NEVER store passwords/secrets in ticket notes or visible UI

CRITICAL: #7 Async/Await Rule - ZERO TOLERANCE

• Every async function MUST contain at least one await expression
• If no await is needed, REMOVE the async keyword
• Lint error: "Async function has no 'await' expression"
• This is a MANDATORY code quality requirement

After generation:

1. Run fdk validate to catch all errors
2. Fix all validation errors before presenting code
3. Check code coverage (minimum 80% required for marketplace)
4. Verify all mandatory files exist

Error Categories & Fixes

For comprehensive error catalog with examples and fixes:

• Load: references/errors/error-catalog.md
• Also see: references/errors/manifest-errors.md, references/errors/oauth-errors.md, references/errors/request-template-errors.md

Top 5 Most Common Errors:

1. Missing app/styles/images/icon.svg - Frontend apps must have icon
2. JSON multiple top-level objects - Merge into single object with commas
3. Host with path/encoded chars - Use FQDN only + <%= context.variable %>
4. Async without await - Add await OR remove async
5. Helper before exports - Move helper functions AFTER exports block

UNIVERSAL ERROR PREVENTION CHECKLIST

BEFORE generating ANY app code, verify ALL of these:

Mandatory Files (Frontend Apps)

• app/styles/images/icon.svg - MUST EXIST - #1 validation failure cause
• app/index.html - MUST include Crayons CDN
• app/scripts/app.js - MUST use IIFE pattern
• app/styles/style.css - MUST exist
• manifest.json - MUST be Platform 3.0 structure
• config/iparams.json - MUST exist (can be empty {})

Request Templates (FQDN Enforcement)

• Host is FQDN only - NO path, NO encoded characters
• Path starts with / - MUST begin with forward slash
• Use <%= context.variable %> - NEVER {{variable}}
• Use <%= iparam.name %> - For app-specific iparams
• Use <%= access_token %> - For OAuth authorization
• All request templates declared in manifest - modules.common.requests

OAuth Structure (If OAuth is used)

• oauth_iparams in oauth_config.json - NOT in regular iparams.json
• Use <%= oauth_iparams.client_id %> - Correct syntax
• options.oauth in request templates - MUST be present
• OAuth config has integrations wrapper - Platform 3.0 requirement

Code Quality

• Helper functions AFTER exports block - FDK parser requirement
• Async functions have await - Or remove async keyword
• No unused parameters - Remove ENTIRELY (not _args)
• Function complexity ≤ 7 - Extract helpers if needed
• IIFE pattern for async initialization - Prevent race conditions

🔒 Security (MANDATORY)

• No command/code execution with user input - No executeCommand(args), eval(args.script)
• SMI args validated - Type check, sanitize, use allowlists
• No logging iparams - No console.log(args.iparams) or console.log(args)
• Safe DOM updates - Use textContent or sanitize before innerHTML
• No secrets in notes - No passwords/tokens in ticket notes or comments
• Sanitized error logging - Log e.message only, not full error objects

Manifest Structure

• All SMI functions declared in manifest - modules.common.functions
• LOCATION PLACEMENT VERIFIED - MANDATORY PRE-GENERATION CHECK:
  • ✅ full_page_app → MUST be in modules.common.location
  • ✅ cti_global_sidebar → MUST be in modules.common.location
  • ✅ ticket_sidebar → MUST be in modules.support_ticket.location (NOT common)
  • ✅ contact_sidebar → MUST be in modules.support_contact.location (NOT common)
  • ✅ asset_sidebar → MUST be in modules.service_asset.location (NOT common)
  • ❌ NEVER put full_page_app in product modules
  • ❌ NEVER put product locations in common module
• At least one product module - Even if empty {}
• No Platform 2.x patterns - No whitelisted-domains, no product
• No scheduled events in manifest - Create dynamically with $schedule.create()

UI Components (Frontend Only)

• Crayons components (not plain HTML) - NO <button>, <input>, etc.
• Crayons CDN included - BOTH script tags (ESM and nomodule)
• Use fwClick, fwInput events - Not click, input

JSON Structure Validation (Pre-Finalization)

• config/requests.json - Single top-level object, all requests as properties ✅
• config/iparams.json - Single top-level object, all iparams as properties ✅
• config/oauth_config.json - Single top-level object with integrations property ✅
• manifest.json - Single top-level object ✅
• No multiple top-level objects ✅ - Merge if found
• Proper comma placement ✅ - Commas between properties, no trailing commas
• Valid JSON syntax ✅ - Run fdk validate to verify

Autofix Process:

1. Run fdk validate to identify JSON errors
2. Fix multiple top-level objects by merging into single object
3. Fix comma placement (add missing, remove trailing)
4. Re-run fdk validate until it passes
5. Only finalize when validation passes completely

Reference: See .cursor/rules/validation-autofix.mdc for detailed autofix patterns.

IF ANY ITEM FAILS → STOP AND FIX BEFORE PROCEEDING

Pre-Finalization Validation & Autofix

CRITICAL: Fix ALL errors - Platform errors AND Lint errors. ZERO TOLERANCE.

After creating ALL app files, you MUST AUTOMATICALLY:

1. Run fdk validate - AUTOMATICALLY run validation (DO NOT ask user)
2. Parse ALL errors - Both platform errors AND lint errors need fixing
3. Attempt Auto-Fix (Iteration 1 - ALL Errors):
   • Fix JSON structure errors (multiple top-level objects)
   • Fix comma placement (missing/trailing commas)
   • Fix template syntax ({{variable}} → <%= variable %>)
   • Create missing mandatory files (icon.svg, iparams.json)
   • Fix FQDN issues (host with path → FQDN only)
   • Fix path issues (missing / prefix)
   • Fix lint: Remove async if no await, remove unused params
   • Re-run fdk validate
4. Attempt Auto-Fix (Iteration 2-6 - Keep iterating):
   • Fix manifest structure issues
   • Fix request template declarations
   • Fix function declarations
   • Fix OAuth structure (if applicable)
   • Fix location placement
   • Fix lint: Extract helpers for complexity > 7
   • Re-run fdk validate
5. After iterations:
   • ✅ If ALL errors (platform AND lint) are resolved → Present app as complete
   • ⚠️ If ANY errors persist → Keep fixing, NEVER say "complete"

What to FIX (Platform Errors) - BLOCKING:

• ✅ JSON parsing errors
• ✅ Missing required files
• ✅ Manifest structure errors
• ✅ Request template errors (FQDN, path, schema)
• ✅ Missing declarations in manifest
• ✅ OAuth structure errors
• ✅ Location placement errors
• ✅ "name" field in manifest.json → REMOVE IT

What to FIX (Lint Errors) - ALSO BLOCKING:

• ✅ Async without await → Remove async OR add await
• ✅ Unused parameters → Remove parameter ENTIRELY
• ✅ Function complexity > 7 → Extract helpers
• ✅ Unreachable code → Remove dead code

CRITICAL: You MUST fix ALL errors (platform AND lint). Keep iterating (up to 6 times) until ZERO errors. An app with ANY errors is NOT complete.

Reference: See validation-autofix.mdc for detailed autofix patterns and examples.

Common JSON Structure Errors & Fixes

Error: "Unexpected token { in JSON"

• Cause: Multiple top-level JSON objects
• Fix: Merge into single object with proper commas

Example Fix (requests.json):

// WRONG - Multiple top-level objects
{  "request1": { ... } }
{  "request2": { ... } }

// CORRECT - Single object
{
  "request1": { ... },
  "request2": { ... }
}

Example Fix (iparams.json):

// WRONG - Multiple top-level objects
{  "param1": { ... } }
{  "param2": { ... } }

// CORRECT - Single object
{
  "param1": { ... },
  "param2": { ... }
}

App Completion Gates - MANDATORY

🚨 ZERO TOLERANCE: An app is NEVER complete unless ALL gates pass.

Gate 1: Mandatory Files Exist

• manifest.json exists - APP CANNOT EXIST WITHOUT THIS
• config/iparams.json exists (can be empty {})
• For frontend apps: app/index.html, app/scripts/app.js, app/styles/images/icon.svg
• For serverless apps: server/server.js

Gate 2: Manifest-to-File Consistency (NEW - CRITICAL)

• Every file referenced in manifest MUST exist:
  • If location.*.url: "index.html" → app/index.html MUST exist
  • If location.*.icon: "styles/images/icon.svg" → app/styles/images/icon.svg MUST exist
  • If modules.*.events exists → server/server.js MUST exist with handlers
  • If modules.common.functions exists → server/server.js MUST exist with functions
• Every template in requests.json MUST be in manifest:
  • For EACH key in config/requests.json, there MUST be a matching key in modules.common.requests

Gate 3: Manifest Structure Valid

• "platform-version": "3.0"
• No empty blocks (functions: {}, requests: {}, events: {})
• All declared functions have implementations in server.js
• All declared requests exist in config/requests.json

Gate 4: OAuth Complete (if used)

• Every integration has display_name
• Every integration has token_type
• Every oauth_iparam field has description

Gate 5: Code Quality (NEW - CRITICAL)

• No function with complexity > 7 - Refactor before finalizing
• No async without await - Add await or remove async
• No unused variables - Remove or prefix with _

Gate 6: Validation Passes

• fdk validate returns 0 platform errors
• fdk validate returns 0 lint errors
• Warnings reviewed and addressed if critical

IF ANY GATE FAILS:

• ❌ DO NOT report app as "complete" or "generated successfully"
• ❌ DO NOT proceed to "Next Steps"
• ✅ Fix the issue and re-validate
• ✅ Only report success when ALL gates pass

Post-Generation Message

After successfully generating an app, provide a concise summary:

✅ App generated successfully in <app-directory>/

Validation: [0 platform errors, 0 lint errors]

Next steps:
1. cd <app-directory>
2. fdk run
3. Test in Freshworks product with ?dev=true

DO NOT automatically generate:

• ❌ Detailed validation reports (.validation-report.md)
• ❌ Apps summary documents (APPS-SUMMARY.md)
• ❌ Extensive feature lists or comparisons
• ❌ Long "Next Steps" sections with multiple subsections

Only generate these when user explicitly requests:

• ✅ "Create a validation report"
• ✅ "Generate a summary document"
• ✅ "Write detailed documentation"
• ✅ "Compare the apps"

Keep post-generation output minimal and focused on immediate next steps.

Documentation Generation Rules

Mandatory Files (ALWAYS create)

• ✅ manifest.json - App manifest
• ✅ server/server.js - Server code (if serverless/hybrid)
• ✅ app/ files - Frontend code (if frontend/hybrid)
• ✅ config/ files - Configuration (iparams, requests, oauth)
• ✅ README.md - Basic installation and usage guide

Optional Files (ONLY create when explicitly requested)

• ❌ .validation-report.md - Detailed validation report
• ❌ APPS-SUMMARY.md - Multi-app comparison
• ❌ ARCHITECTURE.md - Technical architecture docs
• ❌ CHANGELOG.md - Version history
• ❌ .gitignore - Git ignore file
• ❌ package.json - NPM package file (not needed for FDK apps)

User must explicitly say:

• "Create a validation report"
• "Generate a summary document"
• "Write architecture documentation"
• "Add a changelog"

Default behavior: Create only mandatory files + basic README.md

Installation

Installing a Skill (works across all tools)

Install from this marketplace using the Agent Skills standard:

npx @anthropic-ai/add-skill https://github.com/freshworks-developers/freshworks-platform3/tree/main/skills/app-dev

Installing in Cursor

npx skills add https://github.com/freshworks-developers/freshworks-platform3 --skill freshworks-app-dev-skill

Installing in Claude Code

# Install a full plugin
claude plugin install <plugin-path>

# Or add individual skills
npx @anthropic-ai/add-skill https://github.com/freshworks-developers/freshworks-platform3/tree/main/skills/app-dev

Test-Driven Validation

Use these references to validate generated apps:

Golden Tests (Correct Patterns)

references/tests/golden.json - 4 test cases:

1. Minimal Frontend App
2. Serverless App with Events
3. Hybrid App with SMI and External API
4. OAuth Integration

Usage: Generated apps should match these structural patterns.

Refusal Tests (Invalid Patterns)

references/tests/refusal.json - 8 test cases:

1. Platform 2.3 manifest → Reject
2. whitelisted-domains → Reject
3. $request.post() → Reject
4. Plain HTML buttons → Reject
5. Missing engines → Reject
6. OAuth without integrations → Reject
7. Location in wrong module → Reject
8. Missing Crayons CDN → Reject

🔒 Security Refusal Tests (see .cursor/rules/security.mdc for details): 9. Command injection patterns → Reject 10. Code execution patterns → Reject 11. Credential logging → Reject 12. XSS patterns → Reject 13. Secrets in notes → Reject

Usage: Never generate these patterns.

Violation Tests (Common Mistakes)

references/tests/violations.json - 10 test cases:

1. Async without await
2. Unused parameters
3. High complexity
4. Variable scope issues
5. Missing icon.svg
6. Request not declared
7. SMI function not declared
8. OAuth missing options
9. Missing alwaysApply in rules
10. Missing product module

🔒 Security Violation Tests (see .cursor/rules/security.mdc): 11-15. Input validation, logging, XSS, sensitive data violations

Usage: Check generated code against these violations.

Product Module Quick Reference

Supported Modules by Product

Freshdesk Modules:

• support_ticket - Ticket management
• support_contact - Contact management
• support_company - Company management
• support_agent - Agent management
• support_email - Email management
• support_portal - Portal management

Freshservice Modules:

• service_ticket - Service ticket management
• service_asset - Asset management
• service_change - Change management
• service_user - User/Requester management

Freshsales Modules:

• deal - Deal management
• contact - Contact management
• account (or sales_account) - Account management
• lead - Lead management
• appointment - Appointment management
• task - Task management
• product - Product management
• cpq_document - CPQ document management
• phone - Phone management

Freshcaller Modules:

• call - Call management
• caller_agent - Agent management
• notification - Notification management

Freshchat Modules:

• chat_conversation - Conversation management
• chat_user - User management

Location Placements

Common Locations (configured at modules.common.location):

• full_page_app - Full page application
• cti_global_sidebar - CTI global sidebar (Freshdesk/Freshservice only)

Freshdesk support_ticket Locations (configured at modules.support_ticket.location):

• ticket_sidebar - Ticket sidebar
• ticket_requester_info - Requester info section
• ticket_top_navigation - Top navigation bar
• ticket_background - Background app
• time_entry_background - Time entry background
• ticket_attachment - Ticket attachment section
• ticket_conversation_editor - Conversation editor
• new_ticket_requester_info - New ticket requester info
• new_ticket_background - New ticket background

Freshservice service_ticket Locations (configured at modules.service_ticket.location):

• ticket_sidebar - Ticket sidebar
• ticket_requester_info - Requester info section
• ticket_conversation_editor - Conversation editor
• ticket_top_navigation - Top navigation bar
• ticket_background - Background app
• new_ticket_background - New ticket background
• new_ticket_sidebar - New ticket sidebar
• new_ticket_description_editor - New ticket description editor

Freshservice service_asset Locations (configured at modules.service_asset.location):

• asset_top_navigation - Asset top navigation
• asset_sidebar - Asset sidebar

Freshservice service_change Locations (configured at modules.service_change.location):

• change_sidebar - Change sidebar

Location Placement Rules:

• full_page_app, cti_global_sidebar → modules.common.location
• All product-specific locations → modules.<product_module>.location

Module-to-User-Intent Mapping

Constraints (Enforced Automatically)

• Strict mode: Always reject Platform 2.x patterns
• No inference without source: If not in references, respond "Insufficient platform certainty"
• Terminal logs backend only: console.log only in server/server.js, not frontend
• Production-ready only: Generate complete, deployable apps
• Forbidden patterns: Listed in refusal tests
• Required patterns: Listed in golden tests

Serverless Events Reference

For complete event list by product:

• Load: references/events/event-reference.md

Key events:

• onAppInstall (MUST include if app uses iparams)
• onAppUninstall (MUST include if app has scheduled events/webhooks)
• onTicketCreate, onTicketUpdate (in product modules)
• Scheduled events created dynamically with $schedule.create() - NOT declared in manifest

Request Templates & OAuth

For detailed request template syntax and OAuth configuration:

• Load: references/architecture/request-templates-latest.md
• Load: references/architecture/oauth-configuration-latest.md
• Load: references/api/request-method-docs.md

Quick Rules:

• Host must be FQDN only (no path)
• Path must start with /
• Use <%= context.variable %> for iparams
• Use <%= access_token %> for OAuth
• OAuth requests need "options": { "oauth": "integration_name" }

Jobs Feature

For Jobs documentation:

• Load: references/runtime/jobs-docs.md

Quick pattern:

1. Declare in manifest: modules.common.jobs.jobName
2. Invoke from frontend: client.jobs.invoke("jobName", "tag", {data})
3. Handle in server: exports.jobName = async function(args) { ... }

Summary

This skill provides:

• 140+ reference files for progressive disclosure
• 3 Cursor rules (auto-installed to user's project)
• App templates (frontend, serverless skeletons)
• Test patterns (golden, refusal, violation cases)
• Installation automation (rules-only install)
• Comprehensive module, location, and event references
• Request template and OAuth integration patterns
• Jobs feature documentation

When uncertain about any Platform 3.0 behavior, load the relevant reference file from references/ before proceeding.