> ## Documentation Index
> Fetch the complete documentation index at: https://support.metaview.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Day-One Brief

> Turn a signed candidate's interview history into onboarding context for their new manager — and a checklist of what was promised during hiring.

The day someone signs, their interviews stop being evaluation material and become onboarding context — then usually go unread. The manager starts from a blank page while a record of what this person wants, how they work, and what they were told sits in the hiring archive.

This skill reads that record and produces a brief for the first 1:1, plus a list of the expectations set during the process.

<Info>
  **Requires:** the AI Notetaker (interview history). A connected ATS adds profile and application context.

  Deliberately excludes compensation and interviewer evaluations — see below.
</Info>

## What it does

Two passes over the candidate's interviews:

<Columns cols={2}>
  <Card title="Candidate-side" icon="user">
    Their own words: why they said yes, how they say they work best, what they want to grow into, and the concerns they raised.
  </Card>

  <Card title="Company-side" icon="building">
    What *we* said: scope and ownership commitments, first-project framing, team growth plans, flexibility, growth path — with who said it and in which round.
  </Card>
</Columns>

<Tip>
  Early attrition is often a hiring promise nobody remembers making. The company-side pass is what turns this from a nice-to-have into a retention tool.
</Tip>

## What this brief deliberately excludes

<Warning>
  This is onboarding material, not the hiring record. The skill enforces four red lines:

  * **No interviewer evaluations.** Scorecard verdicts and “we almost didn't…” content never appear. This person was hired; the evaluation is over.
  * **No compensation.** Comp expectations, negotiation history, and package details stay between the candidate, recruiting, and HR.
  * **No sensitive disclosures.** Health, family circumstances, visa stress beyond start-date logistics, and vented frustrations about named people are filtered out.
  * **Self-identified growth areas only.** If you ask for “weaknesses to watch”, it includes only what the candidate said about themselves, framed as ramp support.
</Warning>

## What you'll need

Three anchors — the skill will ask rather than guess:

1. **Who signed** (full name)
2. **Who they'll report to**
3. **Their start date**

Optionally: the role's JD, and anything the manager specifically wants to know.

## How to use it

<AccordionGroup>
  <Accordion title="Standard brief">
    “Maya Chen signed for the Platform PM role, starts Sept 1, reporting to Dan — build the day-one brief.”
  </Accordion>

  <Accordion title="Just the promises">
    “What did we actually promise Tomás during his process? His manager wants the list.”
  </Accordion>

  <Accordion title="First 1:1 prep">
    “First 1:1 prep for our new AE — what should their manager open with?”
  </Accordion>

  <Accordion title="30/60/90 check-in">
    “It's been 60 days since Priya started — build the check-in agenda from what we sold her.”
  </Accordion>
</AccordionGroup>

## What you get

A one-page brief with:

* **Who you're getting** — the arc of their background in 3–4 bullets
* **What made them say yes** — pull factors in their words, ranked by how strongly they came up
* **How they say they work best** — autonomy, feedback style, communication rhythm, working pattern
* **What they want to grow into** — the ambitions they voiced
* **What we promised** — each commitment, who said it, which round. The checklist that prevents week-three regret
* **Concerns they raised** — each with a suggested proactive move
* **First 1:1 starters** — 5–6 openers grounded in the interviews
* **First two weeks** — suggested actions mapped to the above

## Safeguards built into this skill

* Everything is traceable to a specific conversation. Where it's unclear whether something was a promise or a musing, the skill quotes it and lets you judge.
* The brief goes to whoever requested it. Sending it on to the manager is your call — the skill offers a draft message rather than sending one.
* On a 30/60/90 check-in, it's explicit that it cannot observe post-hire reality. It structures the conversation; the manager supplies the truth.

## Install this skill

Copy the following into a `SKILL.md` file and add it to your AI client. See [installation steps](/ai-skills/overview#install-a-skill).

````markdown SKILL.md expandable theme={null}
---
name: metaview-day-one-brief
description: >
  Turn a signed candidate's interview history into an onboarding brief for their new manager.
  Trigger on "X just signed / accepted — build the day-one brief", "prep [manager] for their
  new hire", "what did we promise X during the process", "onboarding brief for our new AE",
  "first 1:1 prep for the new starter", or 30/60/90 check-ins like "are we delivering what we
  sold X". Compiles what the new hire said about their motivations, working style, growth
  goals, and concerns — plus every non-comp promise our side made — into a brief the manager
  reads before the first 1:1. Deliberately excludes compensation and interviewer evaluations.
---

# Metaview Day-One Brief

The day someone signs, their interviews stop being evaluation data and become onboarding gold — then evaporate. The manager starts the relationship from a blank page while a rich record of what this person wants, how they work, and what they were promised sits unread in the hiring archive. This skill reads it and hands the manager a brief for the first 1:1 — and a checklist of the expectations set during hiring, so the company delivers what it sold.

## Why this is powerful

Early attrition is usually a broken hiring promise nobody remembers making — "you'll own this end to end" said in round two, forgotten by week three. Those promises, and everything the new hire said about what they're running toward, are already on tape. This is a retention play built from recruiting data: the manager's first 1:1 starts from "you mentioned wanting X — here's how we'll get you there" instead of "so, tell me about yourself." It works because the process was already recorded.

## Red lines (what this brief is and isn't)

- **Candidate's own words only** for who-they-are sections. Interviewer opinions, scorecard verdicts, and "we almost didn't…" content are **never included** — this person was hired; the evaluation is over. The one inversion: **our side's statements** are mined for the *promises* section (what we committed to or sold).
- **No compensation.** Comp expectations, negotiation history, and package details stay out by default — that's between the candidate, recruiting, and HR. If comp came up heavily, one neutral line ("comp and level details live with recruiting") is the maximum. Include actual figures only if the user explicitly asks and says why.
- **Filter sensitive disclosures.** Health, family circumstances, visa stress beyond start-date logistics, and vented frustrations about named people or employers don't go in. Themes are fine ("wanted more autonomy than their last role offered"); grievances verbatim are not.
- **If the user asks for "weaknesses to watch":** include only growth areas the candidate *said about themselves* ("I want to get better at X"), framed as ramp support. Interviewer-perceived weaknesses belong in the hiring record — point them to metaview-interview-debrief for that, and keep it out of this document.
- The brief is delivered to the requester. Sending it onward to the manager is their call — offer a draft message, don't send unprompted.

## Workflow

### Step 1 — Setup

`get_user_context` once per session. Then ask for the three anchors (don't guess): **who signed** (full name), **who they'll report to**, and **the start date**. Optional but sharpening: the role's JD and anything the manager specifically wants to know.

### Step 2 — Pull the process

1. Resolve the candidate: `list_field_values` with `field_id="default:candidate"`, `search_term="<full name>"` (fall back to email). Disambiguate duplicates with the user.
2. All their conversations: `search_conversations` filtered `{"field_id": "default:candidate", "operation": "includes_one_of", "value": ["<id>"]}`. A typical hire has 2-6 rounds: read full transcripts for up to ~4 (one hour ≈ 10-15K tokens), summaries plus the two most substantive transcripts beyond that. Prefer the screen (motivations live there) and the HM round (promises live there).
3. `fetch_candidates` for the profile (background arc for the snapshot). **Do not use the scorecards section** — evaluations are out of scope by design.
4. If this was a rehire or they had an earlier process, use the latest process; mention earlier contact only as warm context ("second time around — they stayed interested for a year").

### Step 3 — Extract, two-sided

Walk the transcripts twice with different filters:

- **Candidate-side pass** (their statements only): why they said yes (pull factors, in their words) · what they said they wanted to move away from, in their words · how they say they work best (autonomy, feedback style, communication, rhythm) · what they want to grow into (direction, skills, the ambition they voiced) · concerns and hesitations they raised · texture for conversation starters (the side project, the story they lit up telling, the question they asked twice).
- **Company-side pass** (our statements only): every commitment or strong impression we gave — scope and ownership ("you'll own X end to end"), first-project framing, team growth plans, flexibility and working model, growth path, "we'll fix Y by the time you join". Note who said it and in which round. These are the expectations the manager inherits, whether or not they were in the room.

Everything traceable to a specific conversation; when in doubt whether something was a promise or a musing, quote it and let the reader judge.

### Step 4 — Compose the brief

```
# Day-one brief — <Name> · <Role> · starts <date> · for <manager>
Built from <n> interviews (<dates>). Their own words throughout; no comp, no interview
evaluations — this is onboarding material, not the hiring record.

## Who you're getting (30-second read)
3-4 bullets: the arc of their background and the work they said they're proudest of.

## What made them say yes
The pull factors, in their words, ranked by how often/strongly they came up. These are the
expectations to honor in the first month.

## How they say they work best
Stated working style: autonomy vs direction, feedback preferences, communication rhythm,
office/remote pattern. Flag anything that differs from how the team currently runs.

## What they want to grow into
Direction and ambitions they voiced, plus skills they said they want to build.

## What we promised (deliver on this)
Each commitment: what was said (quote or close paraphrase) · who said it · which round.
This is the checklist that prevents week-three regret.

## Concerns they raised — defuse early
Hesitations voiced during the process, each with a suggested proactive move.

## First 1:1 starters
5-6 personalized openers grounded in the interviews — warm, specific, zero small-talk.

## First two weeks — suggested moves
3-5 actions mapped to the above (e.g. "give them the X migration early — ownership of
exactly that is why they joined").
```

Keep it to one page of reading. Every section earns its place or gets cut — a thin process produces a short brief, and that's fine.

### Step 5 — Deliver and offer follow-ups

Deliver in chat first for the requester's review. Then offer: a document version (for the manager or the onboarding wiki) · a draft DM to the manager (send only on confirmation) · and a **30/60/90 check-in**: on request (or scheduled, if the surface supports it), regenerate the promises + concerns sections as a check-in agenda — "what we said vs what's happening" prompts for the manager's 1:1. Be clear it can't observe post-hire reality (1:1s aren't in Metaview); it structures the conversation, the manager supplies the truth.

## Edge cases

- **Only one recorded round:** say the brief is built from a single screen; lean on the profile for the snapshot and keep the promises section honest about coverage ("HM round wasn't recorded — check with <interviewer> for commitments made there").
- **Unrecorded rounds in the middle:** list them in a coverage line so missing promises aren't assumed absent.
- **Manager was in the interviews:** the brief shifts from "meet your hire" to "here's the consolidated record" — the promises checklist and starters still carry it.
- **Internal transfer:** works, but interviews may be sparse; blend with what the user tells you about their current team.
- **The candidate hasn't actually signed** (user jumping the gun): confirm before building — this skill is for after the yes; before it, use metaview-interview-debrief (decision support) or the offer-positioning section there.

## Example invocations

- "Maya Chen signed for the Platform PM role, starts Sept 1, reporting to Dan — build the day-one brief."
- "What did we actually promise Tomás during his process? His manager wants the list."
- "First 1:1 prep for our new AE — what should their manager open with?"
- "It's been 60 days since Priya started — build the check-in agenda from what we sold her."
````

## Related

<Columns cols={2}>
  <Card title="Interview Debrief" icon="messages" href="/ai-skills/interview-debrief">
    The hiring record — including interviewer evaluations this brief excludes.
  </Card>

  <Card title="Sharing your notes" icon="share-nodes" href="/using-ai-notes/notes-properties/sharing">
    Controls for who can see a conversation and its notes.
  </Card>
</Columns>
