gate-exchange-activitycenter

Activity center aggregates platform campaigns (trading competitions, airdrops, newcomer activities, referral activities, etc.), supporting activity recommendations and my activities entry.

General Rules

⚠️ STOP — You MUST read and strictly follow the shared runtime rules before proceeding. Do NOT select or call any tool until all rules are read. These rules have the highest priority. → Read ./references/gate-runtime-rules.md

• Only call MCP tools explicitly listed in this skill. Tools not documented here must NOT be called, even if they exist in the MCP server.

Trigger Scenarios: User mentions "recommend activities", "what activities", "airdrop activities", "trading competition", "VIP activities", "my activities", etc.

⚠️ MANDATORY RESPONSE FORMAT: When this skill is triggered, AI responses MUST strictly follow the Response Templates defined in this document. Free-form responses are FORBIDDEN.

MCP Dependencies

Required MCP Servers

MCP Tools Used

Query Operations (Read-only)

• cex_activity_get_my_activity_entry
• cex_activity_list_activities
• cex_activity_list_activity_types

Authentication

• Credentials Source: Local Gate MCP deployment (GATE_API_KEY, GATE_API_SECRET)
• API Key Required: Yes
• Permissions: Activity:Read
• Never ask the user to paste secrets into chat; rely on the configured MCP session only.
• API Key Provisioning Reference: https://www.gate.com/myaccount/profile/api-key/manage (create or rotate keys outside the chat when the local MCP setup requires them).

Installation Check

• Required: Gate (main)
• Install: Use the local Gate MCP installation flow for the current host IDE before continuing.
• Continue only after the Gate MCP session is configured with the credentials listed above; do not switch to browser auth or ask the user to paste secrets into chat.

MCP Mode

Read and strictly follow references/mcp.md, then execute this skill's activity-center workflow.

• SKILL.md keeps routing and recommendation logic.
• references/mcp.md is the authoritative MCP execution layer for entry/list/type queries, filter handling, and degraded output behavior.

Routing Rules

MCP Tools

API Parameter Constraints

⚠️ CRITICAL: The page_size parameter is strictly limited to a maximum of 3. Any value greater than 3 is FORBIDDEN.

Scenario 1: Activity Recommendation

1.1 Hot Recommendation (No Type Specified)

Trigger: "recommend activities" "what activities" "recent activities"

API Call:

cex_activity_list_activities?recommend_type=hot&sort_by=default&page=1&page_size=3

Rules:

• ✅ Do NOT ask for type clarification, call API directly
• ✅ Fixed page_size=3 (maximum allowed), no pagination

1.2 Type-Based Recommendation

Trigger: "airdrop activities" "trading competition" "VIP activities" "earn activities"

API Call Sequence:

1. cex_activity_list_activity_types → Get type list, match user's type name (case-insensitive)
2. cex_activity_list_activities?recommend_type=type&type_ids=matched_id1,matched_id2&sort_by=time&page_size=3

Keyword Extraction Examples:

• "what airdrop activities are there" → Extract: airdrop
• "VIP activities" → Extract: VIP
• "I'm a VIP, any exclusive activities" → Extract: VIP
• "any low-risk activities" → Extract: earn

1.3 Scenario-Based Recommendation

Trigger: "I have GT, what activities can I join" "futures activities"

API Call:

• If mappable to type: Use recommend_type=type with type_ids
• If scenario/asset: Use recommend_type=scenario with keywords

cex_activity_list_activities?recommend_type=scenario&keywords=GT&page_size=3

1.4 Search by Name

Trigger: "help me find position airdrop" "where is VIP exclusive activity" "find test activity"

API Call:

cex_activity_list_activities?recommend_type=scenario&keywords=test_activity&page=1&page_size=3

Parameter Notes:

• Use recommend_type=scenario with keywords for name search
• keywords is the activity name or partial name to search
• ⚠️ Language Conversion: If user input contains non-English text, translate keywords to English before passing to the keywords parameter

Response Template (when results found):

Based on your keywords, I found some possible activities. If you didn't find the activity you're looking for, you can provide a more complete activity name, or click more activities to explore more hot activities.

Scenario 2: My Activities

Trigger: "what activities I joined" "my activities" "enrolled activities"

API Call:

cex_activity_get_my_activity_entry

Rules:

• ✅ Only show entry card, guide user to click and jump
• ✅ Apply URL Processing Rules (see above) to the url field
• ❌ Do NOT filter activity list in this API

Decision Logic

URL Processing Rules

All activity URLs returned by API must be processed before displaying to users:

Why relative paths? API returns relative paths for flexibility:

• Multi-language support: Frontend can dynamically add language prefix (/zh/, /en/, /ja/)
• Multi-environment deployment: Different domains for test/staging/production
• Client flexibility: Web, App, H5 can handle links differently

Display Format: Always use clickable Markdown link: [{activity_title}]({processed_url})

Activity Response Field Specification

For all activity recommendation scenarios (1.1, 1.2, 1.3, 1.4), the response MUST only include the following fields:

Response Format Example:

Rules:

• ✅ Only display the 3 fields listed above
• ✅ Use type_name for activity type (NOT type_id)
• ✅ Process URLs according to URL Processing Rules
• ❌ Do NOT display other fields like id, type_id, hot, img, competition_name, start_at, end_at, etc.

Response Templates

⚠️ MANDATORY: AI responses MUST strictly follow these templates. Free-form or custom responses are FORBIDDEN.

Template Compliance Rules

1. REQUIRED: Every response MUST use the exact template structure defined below
2. REQUIRED: Include the intro text + activity table + follow-up prompt
3. FORBIDDEN: Adding extra commentary, explanations, or custom formatting
4. FORBIDDEN: Omitting any part of the template structure

Activity Recommendation (Has Results)

Required Response Structure:

{Intro Text from table above}

| Activity Title | Activity Type | Activity Link |
|----------------|---------------|---------------|
| {master_one_line} | {type_name} | [View Details]({processed_url}) |
| ... | ... | ... |

{Follow-up Prompt - see below}

Follow-up Prompts by Scenario:

My Activities Template (Scenario 2)

Required Response Structure:

Click [My Events]({processed_url}) to view your enrolled and ongoing activities.

In "My Events" you can:
- View activity progress
- Claim activity rewards
- Track activity status

No Results Template

Required Response Structure:

Currently no [type/keyword] activities are in progress. You can:
- Check the activity center regularly, new activities will be displayed as soon as they launch
- Tell me other activity types you're interested in (e.g., trading, earn, VIP exclusive), and I can filter for you

Error Handling

Safety Rules

1. No investment advice: Activity rewards are platform benefits, not investment guidance
2. No data fabrication: Only show data returned by backend
3. Compliance check: Do not show activities unavailable in user's region
4. No internal exposure: Never mention technical details to users, including:
   • API names (e.g., cex_activity_list_activities)
   • Parameter names (e.g., recommend_type, type_ids)
   • Internal IDs (e.g., type_id=34, id=1499)
   • Error codes or technical error messages
   • Example of BAD response: "No Alpha type activities found (type_id=34)"
   • Example of GOOD response: "Currently no Alpha activities are in progress"

Data Integrity Rules

⚠️ MANDATORY: These rules MUST be followed for ALL data processing and display.

1. Never Guess Data

• ❌ FORBIDDEN: Guessing, estimating, or fabricating any data values
• ❌ FORBIDDEN: Making assumptions about timestamps, numbers, or any computed values
• ✅ REQUIRED: Only display data exactly as returned by API
• ✅ REQUIRED: If data is missing or unclear, state "Data unavailable" instead of guessing

2. Self-Verification Mechanism

Before displaying any processed data, perform self-check:

3. Proactive Validation (Not Reactive Correction)

• ✅ REQUIRED: Validate data BEFORE presenting to user
• ✅ REQUIRED: Double-check time-sensitive information (timestamps, dates)
• ❌ FORBIDDEN: Wait for user to point out errors
• ❌ FORBIDDEN: Display data first and correct later

4. Error Handling for Data Issues

If data validation fails:

1. Do NOT display the problematic data
2. Show a safe fallback message: "Some data is temporarily unavailable"
3. Log the issue internally (do NOT expose to user)

Cross-Skill Routing

Additional References

For detailed scenarios and examples, see references/scenarios.md