sales-skills/sales-xquik

Xquik Platform Help

Step 1 - Gather context

If references/learnings.md exists, read it first for accumulated platform knowledge.

Ask the user:

1. What do you need help with?

   • A) MCP setup for an AI agent
   • B) REST API workflow design
   • C) SDK or agent-skill installation
   • D) Webhook delivery and verification
   • E) Bulk extraction or export planning
   • F) Monitoring accounts or keywords
   • G) Confirmation-gated publishing actions
   • H) Something else - describe it

2. Which integration path are you using?

   • A) MCP server
   • B) REST API
   • C) Agent skill from Xquik-dev/x-twitter-scraper
   • D) TypeScript or Python SDK
   • E) Webhooks
   • F) Not sure yet

3. What output do you need?

   • A) Mentions or post search
   • B) Follower or profile export
   • C) Media download
   • D) Monitor alerts
   • E) Giveaway draw data
   • F) Draft or publish workflow

If the user's request already provides most of this context, skip directly to Step 2.

Step 2 - Route or answer directly

• Social listening strategy or tool comparison -> /sales-social-listening [question]
• Social media publishing calendars -> /sales-social-media-management [question]
• Buyer-intent signals from social data -> /sales-intent [question]
• Competitive monitoring program design -> /sales-social-listening [question]

Otherwise, answer directly from the platform reference below.

Step 3 - Xquik platform reference

Read references/platform-guide.md for capabilities, integration paths, MCP setup, webhook patterns, SDK choices, and workflow examples.

Answer the user's question using only the relevant section. Do not dump the full reference.

Step 4 - Actionable guidance

Focus on the user's exact workflow:

1. MCP setup - verify transport, endpoint, API-key auth, and a small read test
2. REST workflow - pick endpoints, pagination, retries, output shape, and rate-limit handling
3. Bulk extraction - choose the extraction job, destination, polling, and export format
4. Webhooks - verify HMAC handling, retry behavior, idempotency, and HTTPS delivery
5. Publishing - keep write actions explicit and confirmation-gated

If you discover a gotcha or tip not covered in references/learnings.md, append it there.

Gotchas

Best-effort from public docs - review these before acting, especially setup and endpoint details that may change.

• Choose the integration path first. MCP is best for agent workflows, REST is best for applications and pipelines, SDKs are best for typed application code, and the installable skill is best when an AI coding agent needs endpoint-selection guidance.
• MCP requires a dashboard API key. The public MCP endpoint is protected; an unauthenticated request should fail. Add API-key auth through the client config supported by your agent.
• Bulk extraction is a workflow, not one request. Plan job creation, polling, pagination, export handling, and webhook notifications instead of assuming one synchronous response.
• Webhooks need HMAC verification. Treat webhook bodies as untrusted until the signature is verified and use idempotency for retries.
• Publishing actions need explicit confirmation. Keep target account, payload, destination, and user approval visible before creating or changing public X/Twitter content.

Related skills

• /sales-social-listening - Social listening strategy - monitoring setup, tool comparison, sentiment, crisis detection, competitive intel
• /sales-octolens - Octolens platform help - developer-first monitoring with MCP server, REST API, and webhooks
• /sales-xpoz - Xpoz platform help - social data API and MCP server for AI agents
• /sales-socialhose - Socialhose platform help - social listening with REST API and webhooks
• /sales-do - Not sure which skill to use? The router matches any sales objective to the right skill. Install: npx skills add sales-skills/sales --skill sales-do

Examples

Example 1: Connect MCP for an agent

User says: "I want Claude Code to query recent mentions from Xquik." Skill does:

1. Reads the platform guide MCP section
2. Confirms the MCP endpoint and API-key auth path
3. Suggests a small read-only test before building a larger workflow
4. Routes endpoint-selection details to the installable Xquik agent skill when needed Result: Working MCP connection with a verified read path

Example 2: Build a follower export pipeline

User says: "I need a weekly follower export pushed into our CRM." Skill does:

1. Chooses REST or SDK based on the user's stack
2. Designs extraction, polling, pagination, and deduplication
3. Adds webhook or scheduled polling for job completion
4. Maps output fields before sending records to the CRM Result: Reliable export workflow with retry-safe delivery

Example 3: Set up a monitored-account alert

User says: "Alert sales when key accounts post about a competitor." Skill does:

1. Defines the monitored account list and keyword rules
2. Recommends REST or monitor workflows for the alert source
3. Designs webhook verification and Slack or CRM routing
4. Adds a review step for false positives and query tuning Result: Focused account-monitoring workflow with auditable delivery

Troubleshooting

MCP returns unauthorized

Symptom: The MCP endpoint connects but requests fail with an auth error. Cause: Missing, expired, or incorrectly configured dashboard API key. Solution: Recheck the client config, make sure the Authorization bearer header is supported by the agent client, and retry with a small read-only call.

Exports are incomplete

Symptom: Follower, post, or media exports miss expected records. Cause: Pagination, job polling, or filter windows are too narrow. Solution: Read the API reference for the selected extraction route, persist cursors or job IDs, and reconcile counts before sending data downstream.

Webhook events are rejected

Symptom: Your endpoint receives Xquik webhook events but drops them. Cause: Signature verification, timestamp tolerance, or body parsing mismatch. Solution: Verify the HMAC signature against the raw request body before parsing, return 2xx only after durable handling, and make the receiver idempotent.