Metaview × Claude & ChatGPT

A skills library for the Metaview MCP.

Nine skills that turn Claude and ChatGPT into a working member of your team, prepping interviews, drafting scorecards, spotting bias, briefing execs. A prompt gets you one answer; a skill runs the whole job, the same way, every time. Install once, then ask for the outcome.

Browse the library→Install in two minutes
Jump to
/ 01

From first screen to first day

Notetaker-only, every workspace can run these immediately.

01
Notetaker+ATS optional
metaview-interview-prep

Interview prep, done for you

Builds a pre-interview brief from the candidate's full history: what's been asked and answered, what's still untested, contradictions to probe, and a question plan for your round.

Five minutes of reading replaces an hour of prep, and candidates stop repeating their story round after round.

  • Finds every prior conversation, scorecard, and the candidate's profile
  • Maps what's been covered, so nothing gets re-asked
  • Flags contradictions and gaps, then writes a tailored question plan
  • Recurring daily version pairs with your calendar
For · interviewers, hiring managers, recruiters
Download .zipView SKILL.md
02
Notetaker+ATS optional
metaview-interview-debrief

Evidence-based debriefs

Compiles every round for a candidate into a competency-by-competency evidence matrix, where the panel agrees, where it contradicts itself, what nobody tested, and checks scorecards against the transcripts.

The debrief runs on transcript evidence instead of recall, and "not assessed" stops masquerading as "no concerns".

  • Evidence matrix across every round, each claim linked to its conversation
  • Surfaces panel splits and untested competencies before the meeting
  • Offer-positioning if it's a yes; defensible feedback if it's a no
For · hiring managers, recruiters running debriefs
Download .zipView SKILL.md
03
Notetakerpost-signature
metaview-day-one-brief

Day-one manager brief

Compiles a signed hire's interviews into an onboarding brief for their manager: motivations, working style, growth goals, concerns raised, and every non-comp promise made during the process.

The first 1:1 starts informed, and commitments made in round two don't get forgotten by week three.

  • What made them say yes and how they work best, their own words
  • "What we promised" checklist: each commitment, who said it, which round
  • No comp, no interviewer evaluations, onboarding material, not the hiring record
  • 30/60/90 mode: the check-in agenda from what we sold them
For · hiring managers, onboarding owners
Download .zipView SKILL.md
/ 02

Across all your interviews

Reports + AI fields, with the discipline that makes numbers presentable: reuse before create, calibrate before trust, n always visible.

04
NotetakerReports
metaview-interview-intelligence

Ask questions across every interview

Extracts a chosen data point from every interview, decline reasons, competitor mentions, candidate concerns, via calibrated AI fields, then aggregates the results into answers, charts, and saved reports.

Cohort-level answers in minutes, with the n and the calibration audit attached, numbers that survive leadership scrutiny.

  • Designs the AI field, audits it on real transcripts, then aggregates
  • Charts and dashboards in-chat, plus saved Metaview reports
  • Recurring pulse with a sanity gate, an empty dataset sends an alert, not a report
For · TA leads, recruiting ops
Download .zipView SKILL.md
05
NotetakerReports
metaview-interviewer-coaching

Private coaching from real calls

Reviews an interviewer's transcripts against your interview guide, or the bundled Screening Call Guide, and writes a private coaching pack: quoted moments, missed follow-ups, and what to try next call.

Specific, private feedback from calls that actually happened, no shadowing required.

  • Benchmarks rapport, probing, talk time, motivations, and selling; housekeeping is a one-line completeness check
  • Individual packs are private; team views aggregate, never a league table
  • Training guides modelled on your best interviewer's actual calls
For · interviewers (private), TA leads, enablement
Download .zipView SKILL.md
/ 03

Filling the pipeline

The sourcing agent and Application Review, run with a tight loop.

06
SourcingApp Review+ATS
metaview-sourcing-copilot

Daily sourcing digest

Kicks off searches from a JD, sends a daily top-10 to Slack or Teams, records numbered replies ("1, 4 yes · 3 no, too junior") as feedback on the candidates' profiles, and hands winners to your ATS and sequences.

Reviewing costs one reply, every reply calibrates the search, and the loop doesn't die after week one.

  • No-repeats guarantee; pool-empty guard asks the agent for a fresh pack
  • ICP locked with you before big batches arrive
  • Inbound triage: new great-fit applicants in the same daily digest
For · sourcers, recruiters, top-of-funnel owners
Download .zipView SKILL.md
07
Research+Reports optional
metaview-talent-research

Market research with an inside view

Commissions Metaview's research agent for talent maps, comp benchmarks, and target-company lists, then cross-checks the findings against what candidates say in your own interviews.

Market evidence for a hiring decision in about half an hour, sanity-checked against your own data.

  • Talent maps, feeder-company lists, location comparisons, market-event monitors
  • Inside-view cross-check: where candidates actually come from, what they actually ask for
  • Delivers the so-what, not a forwarded report
For · TA leadership, agencies, sourcers scoping hard roles
Download .zipView SKILL.md
08
NotetakerReportsSourcingApp Review+ATS
metaview-role-launch

One intake, whole role set up

Turns one intake call into the complete role setup: role brief, interview plan, stage note templates mirroring your ATS scorecards, sourcing search, inbound Application Review ICP, tracking report, and a weekly pulse.

A coordinated week of setup happens in one sitting, and JD, interviewers, and sourcing can't drift apart, because every artefact derives from the same brief.

  • Captures the hiring bar from the HM's own words, interviews out the gaps
  • One plan, one confirmation, everything traces to the same source of truth
  • Degrades gracefully: builds what your workspace has, says what it skipped
For · recruiters opening a req, agencies launching a search
Download .zipView SKILL.md
Install

Two minutes, one time.

Skills are markdown files the model loads on demand. You never call them by name, Claude or ChatGPT picks the right one from the ask.

  1. 01

    Connect the Metaview MCP

    Settings → Connectors → add https://mcp.metaview.ai/mcp, sign in with Metaview. The model sees exactly what your Metaview role sees.

  2. 02

    Add skills

    Claude: Settings → Capabilities → Skills → upload the skill zip. Claude Code / Cowork: drop the folder into ~/.claude/skills/. ChatGPT: open the SKILL.md and paste it into a Project's instructions or a custom GPT , the markdown runs the same workflow.

  3. 03

    Ask naturally

    No commands to learn.

    › "prep me for my 2pm"
    › "why are candidates declining?"
    › "launch this role"
Principles

Built-in, on every run.

Starting points, deliberately
Works out of the box; edit the SKILL.md to add your stages, rubric language, and tone.
Read-heavy, write-careful
Confirms before anything shared, live, or irreversible (ATS posts, ICP reranks, sequence adds, new AI fields).
Calibrated numbers only
Scale findings are audited against real transcripts, and the calibration ships with the result.
People-safe by default
Cohorts anonymised; coaching private; team views aggregate, never rank named people.
Honest about gaps
"Not discussed" and "n=4, too thin" are answers, not failures.
Humans decide
Skills draft and evidence; hiring decisions stay yours.
FAQ

Common questions.