Lever

@Lever show me all candidates who applied in the last 7 days for our Senior Engineer role

Found 12 candidates who applied to 'Senior Engineer - Backend' between Jan 15-22, 2025. Top applicants include: Sarah Chen (applied Jan 22, 5 years backend experience, referral from current employee), Marcus Rodriguez (applied Jan 21, previously at similar-stage startup, strong system design portfolio), and Aisha Patel (applied Jan 20, open-source contributor to projects we use). All are currently in 'New Lead' stage awaiting initial screen. Would you like me to prioritize any based on specific criteria or move them to the next stage?

This example assumes the MCP can filter candidates by posting and date range. Lever's API requires specific permissions to read candidate data — your API key must have recruiter-level access. The MCP returns structured applicant data; the AI synthesizes it into a ranked summary.

@Lever move all candidates in 'Phone Screen Scheduled' stage to 'Phone Screen Completed' if their interview was more than 3 days ago

Reviewed 8 candidates in 'Phone Screen Scheduled' stage. Moved 5 candidates to 'Phone Screen Completed' (interviews conducted Jan 18-19). Left 3 candidates unchanged (interviews scheduled for Jan 24-25, still upcoming). Updated candidates: Jordan Lee, Priya Sharma, Alex Thompson, Nina Okafor, David Kim. Stage transition logged in each candidate's activity feed with timestamp and reason.

This is a bulk write operation — the MCP changes candidate stage for multiple records. Lever's API logs all stage transitions, so this action is auditable but irreversible through the MCP. Test the logic on a small batch first. Your API key needs write permissions to candidate records.

@Lever draft a summary email for our hiring manager showing this week's pipeline movement for the Product Designer role

Here's a draft summary for the Product Designer role (week of Jan 15-22): 9 new applications received, 4 candidates advanced to portfolio review, 2 moved to final-round interviews (both scheduled for Jan 25), 1 offer extended (awaiting response). Current pipeline: 23 active candidates across all stages. Bottleneck identified: 6 candidates waiting in 'Portfolio Review' stage for 10+ days — recommend prioritizing feedback from design lead. Attached breakdown shows source attribution (5 from LinkedIn, 3 referrals, 1 agency).

The AI synthesizes Lever data into a narrative report — the MCP provides raw pipeline metrics, the AI interprets trends and flags issues. This example shows how @mention prompts can combine data retrieval with reasoning. The draft is illustrative; actual output depends on your account's posting structure and stage names.

When Lever MCP helps hiring managers track candidate flow

A 6-person startup hiring for three roles runs weekly pipeline reviews. The hiring manager needs candidate counts by stage, recent interview feedback, and offer timelines without opening Lever's UI. The Lever MCP can surface this data in a shared Switchy workspace where the team already discusses priorities. The win is speed: pull stage counts and recent notes in one query instead of clicking through tabs. The threshold is complexity—if you need custom reports or multi-stage filtering, Lever's native dashboards are faster to configure. This MCP works when your questions are simple and your team wants recruiting context alongside product or ops discussions in the same workspace.

Using Lever MCP to draft personalized candidate messages

A technical recruiter at a 15-person agency manages 40 active candidates across four clients. Before each outreach call, she needs the candidate's application history, referral source, and previous interview notes. The Lever MCP lets her query this context in Switchy and draft a personalized follow-up without switching tools. The advantage is continuity: she can reference the candidate's GitHub profile (via another MCP) and Lever notes in the same conversation thread. The limit is write operations—if the MCP doesn't support updating candidate stages or scheduling interviews, she still returns to Lever's UI for those actions. This setup wins when research and drafting happen in one place, and execution happens in Lever.

When Lever MCP speeds up post-interview decision-making

A remote 10-person engineering team interviews three candidates in one day. After each loop, the hiring lead collects feedback from four interviewers scattered across time zones. The Lever MCP can pull all feedback into a Switchy thread where the team discusses trade-offs and makes a decision asynchronously. The benefit is consolidation: no one opens Lever separately to read scorecards, and the discussion stays in one thread. The boundary is real-time collaboration—if your team prefers a synchronous Zoom debrief, pulling feedback into Switchy adds a step instead of removing one. This MCP fits when your decision process is async and you want recruiting data in the same workspace as your other team discussions.