meeting-prep（会议准备）

Meeting Prep

Research-powered meeting preparation with attendee intelligence and company context.

User request: $ARGUMENTS

Before running any commands, read references/nimble-playbook.md for Claude Code constraints (no shell state, no &/wait, sub-agent permissions, communication style).

Instructions

Step 0: Preflight

Run the preflight pattern from references/nimble-playbook.md (5 simultaneous Bash calls: date calc, today, CLI check, profile load, index.md load).

From the results:

• CLI missing or API key unset → references/profile-and-onboarding.md, stop
• Profile exists → read ~/.nimble/memory/people/index.md to identify existing person profiles. Load relevant ~/.nimble/memory/people/ files for attendees before — skip redundant searches, surface prior meeting notes. Follow [[path/entity]] cross-references in person files: if an attendee's file links to [[competitors/widgetco]], load that competitor file for richer context (e.g., recent intel from competitor-intel runs). Also check ~/.nimble/memory/companies/ for cached company research. No same-day report check — meeting-prep is per-meeting, not per-day. Users may prep for multiple meetings in one day. Instead, check entity freshness: if a person/company profile was updated within the last 24 hours, offer to reuse it: "I have a recent profile for [Name] from earlier today. Use it, or refresh?"
• No profile → that's fine. Meeting prep doesn't require onboarding. Proceed to Step 1.

Step 1: Gather Meeting Context

Parse the meeting details from $ARGUMENTS or ask the user.

Calendar shortcut: If the user didn't specify attendees and a calendar connector is available — either a calendar MCP tool (look for list_events in the tool list) or the gws CLI (gws calendar +agenda --today) — offer to pull today's meetings so they can pick one. If neither is available, skip this silently.

If clear (e.g., "prep me for my meeting with Alex Kim at WidgetCo tomorrow"):

• Extract: attendee name(s), company, meeting date/time (if given)
• Confirm briefly: "Preparing briefing for your meeting with Alex Kim at WidgetCo..."

If partial (e.g., "prep me for my meeting tomorrow"):

• Ask one clarifying question in plain text: "Who are you meeting with? (names, titles, and company if you have them)"

If just a person (e.g., "research John Smith"):

• Proceed with the person. Try to infer their company from search results.

Extract these fields:

Meeting type detection — if the user doesn't specify, infer from context clues:

If no signal — don't guess "general external." The meeting type gates whether the Value Positioning section is generated, so it's worth one question. Use AskUserQuestion:

What's the goal of this meeting? - Sales / discovery — pitching, demo, exploring fit - Partnership — integration, co-selling, joint venture - Board / investor — board meeting, investor update, fundraising - Interview — evaluating a candidate - General / other — networking, catch-up, or not sure

Map the answer to the meeting type. If the user picks "General / other", treat as general external (no value positioning section).

The meeting type shapes the briefing focus — specifically, it determines whether the Value Positioning section (Step 4.5 + Step 6) is generated. Value positioning activates for: sales/discovery, partnership, board/investor. It is skipped for: interview, internal, general external.

Step 2: WSA Discovery

Discover available WSAs for each attendee's company domain:

nimble agent list --search "{company-domain}" --limit 20

Run one search per unique company simultaneously. Filter for SERP/PDP WSAs, prefer managed_by: "nimble", validate with nimble agent get --template-name {name}. Cache discovered names + params. Pass them to attendee agents in Step 3 for richer data. If no WSAs found, continue with nimble search alone.

Step 3: Per-Attendee Research (sub-agents)

Read references/attendee-agent-prompt.md for the full agent prompt template. Follow the sub-agent spawning rules from references/nimble-playbook.md (bypassPermissions, batch max 4, explicit Bash instruction, fallback on failure).

Check memory first. For each attendee, check ~/.nimble/memory/people/[name-slug].md. If a profile exists and is < 30 days old, load it as known context and pass it to the agent so it focuses on what's new. If > 30 days old, run a full refresh.

Spawn nimble-researcher agents (agents/nimble-researcher.md) with mode: "bypassPermissions". One agent per attendee. Pass discovered WSA names from Step 2 to each agent for enrichment.

Important: The Nimble API has a 10 req/sec rate limit per API key. With each agent running 4-5 searches, limit concurrent agents to 2 per batch. For 3+ attendees, batch in groups of 2.

Call estimation & Scaled Execution: Before launching agents, estimate total API calls: ~5 searches per attendee + ~4 company searches + 3-5 extractions = ~(5 × N) + 9 calls. For 3+ attendees (15+ calls), tell agents to use extract-batch for page extractions instead of individual calls. See the Scaled Execution pattern in references/nimble-playbook.md for tier selection.

Batch 1 (2 agents simultaneously):

• Attendee 1 research
• Attendee 2 research

Batch 2 (if needed):

• Attendee 3 research
• Attendee 4 research

Single attendee optimization: If only one person, run the searches directly from the main context instead of spawning an agent — saves overhead.

Fallback: If any agent fails or returns empty, run those searches directly from the main context. Don't leave gaps in the briefing.

Step 3.5: Gap Check

Before proceeding, verify every attendee has at least a title and company confirmed.

For any attendee with < 3 meaningful results or "Role Unknown":

1. Run a --focus social fallback search directly (this searches social platform people indices and is the most reliable way to find someone): nimble search --query "[Name] [Company]" --focus social --max-results 5 --search-depth lite
2. If --focus social is unavailable, fall back to: nimble search --query "[Name]" --include-domain '["linkedin.com"]' --max-results 5 --search-depth lite
3. Try name variations: "[First] [Last]", "[Full Name] [Company] [Title if known]"

Do NOT present a briefing with "Role Unknown" — exhaust social search first. If still nothing after fallbacks, note it honestly: "Limited public presence — could not confirm role. Consider asking for their LinkedIn URL."

Also collect LinkedIn profile URLs for each attendee during this step if not already found. These are high-value for the final briefing output and Notion distribution.

Step 4: Company Research

Research the attendees' company for meeting-relevant context. This is a lighter version of company-deep-dive — focused on what's useful for the conversation, not a full 360°.

Company name quoting: If the company name contains common words that cause noisy results (e.g., "Acme Supply", "Nova Dynamics", "Global Industries"), wrap it in escaped quotes: "\"Acme Supply\" news". Use --include-domain '["[domain]"]' as an alternative anchor.

Make these Bash calls simultaneously:

• nimble search --query "\"[Company]\" news" --focus news --start-date "[14-days-ago]" --max-results 8 --search-depth lite
• nimble search --query "\"[Company]\" product launch OR announcement" --focus news --start-date "[14-days-ago]" --max-results 5 --search-depth lite
• nimble search --query "about" --include-domain '["[domain]"]' --max-results 3 --search-depth lite
• nimble search --query "\"[Company]\" funding OR raised OR investors" --max-results 5 --search-depth lite

If your user's company profile exists, also run:

• nimble search --query "[Company] [UserCompany] OR [user-domain]" --max-results 5 --search-depth lite

This catches any existing relationship between the two companies — prior partnerships, mentions, shared investors, or competitive overlap.

If < 3 results from the news searches, retry without --start-date.

Date validation: When including company news in the briefing, verify that the event date (when something actually happened) is recent, not just the article date. See references/nimble-playbook.md → "Signal Date Validation" for details. If a snippet uses past-tense language like "last year" or "back in Q3", treat it as background context rather than recent news.

If the company was already researched (exists in ~/.nimble/memory/companies/), load the existing profile and only run the news search for fresh updates.

Step 4.5: Value Positioning Research

Skip this step if the meeting type is interview, internal, or general external.

This step cross-references what you learned about the attendee's company (Step 4) with the user's own business profile to find concrete positioning angles. It works best when business-profile.json exists with at least company.name and company.domain.

If no profile exists, skip searches that reference the user's company or competitors (searches 2, 4, 5) and rely on generic research (searches 1, 3) for positioning insights. Use any WSAs discovered in Step 2 for richer attendee company data. The Value Positioning section will be thinner but still useful — pain-to-solution mapping and tech stack discovery work without a profile.

Load the user's sales context from ~/.nimble/business-profile.json:

• sales_context.key_differentiators — what makes the user's product unique
• sales_context.integration_partners — tools the user's product connects with
• sales_context.case_studies — similar customers and outcomes
• sales_context.common_objections — pre-built objection responses
• competitors — tracked competitors (check if the attendee's company uses any)

If sales_context doesn't exist in the profile, the skill still works — the value positioning section will rely on web research alone rather than profile-enriched data. Mention at the end: "Tip: Add sales context to your profile for richer positioning next time."

Make these Bash calls simultaneously (3-5 searches depending on available data):

1. nimble search --query "\"[AttendeeCompany]\" tech stack OR tools OR platform OR uses" --max-results 5 --search-depth lite → Discover what tools/platforms they use — match against integration_partners
2. nimble search --query "\"[AttendeeCompany]\" [UserCompany] OR [user-domain]" --max-results 5 --search-depth lite → Any existing relationship, mentions, or competitive overlap (skip if already run in Step 3)
3. nimble search --query "\"[AttendeeCompany]\" challenges OR pain points OR struggling OR migrating" --max-results 5 --search-depth lite → Pain signals to map against user's value props
4. (If competitors list exists) nimble search --query "\"[AttendeeCompany]\" [CompetitorName1] OR [CompetitorName2]" --max-results 5 --search-depth lite → Check if they use a competitor — critical for displacement positioning
5. (If case_studies exist with matching industry) nimble search --query "[UserCompany] [attendee-industry] case study OR customer story" --max-results 5 --search-depth lite → Find published case studies in the attendee's industry to reference

From the results, extract:

• Tools/platforms they use (for integration hooks)
• Pain signals or challenges (for value mapping)
• Competitor usage (for displacement angles)
• Industry match to existing case studies (for social proof)

This data feeds directly into the Value Positioning section in Step 5.

Step 5: Deep Extraction

From Steps 3-4.5, identify the top 3-5 most informative URLs across all results. Prioritize:

• Attendee's own LinkedIn posts, articles, or talks
• Recent company announcements directly relevant to the meeting
• Interviews or profiles of the attendee
• The company's about/team page (if attendee title wasn't found)
• (If value positioning active) Pages revealing their tech stack or tool usage
• (If value positioning active) Articles about their challenges or migration plans

Make one Bash call per URL, all simultaneously:

nimble extract --url "https://..." --format markdown

For extraction failures, follow the fallback in references/nimble-playbook.md.

Single attendee + known company: Skip company extraction, focus on person URLs. Multiple attendees: Prioritize person-specific URLs over company-level ones.

Step 6: Synthesize Briefing

Structure the output as a meeting prep briefing. Adapt focus based on meeting type.

# Meeting Prep: [Company Name]
*[Meeting date/time if known] | Prepared [today's date]*

## Quick Take
[2-3 sentences: who you're meeting, why it matters, and the one thing to know
going in. This is the "read nothing else" paragraph.]

## Attendees

### [Name] — [Title]
**Background:** [Current role, time in position, career trajectory highlights]
**Recent Activity:** [What they've been posting, speaking about, or working on.
  Direct quotes from posts/talks when available.]
**Conversation hooks:** [2-3 specific things to reference — shared connections,
  their recent project, a post they wrote, a talk they gave]
**Notes from prior meetings:** [If exists in memory — what was discussed, their
  preferences, open items. "No prior meetings on file" if none.]

[Repeat for each attendee]

## Relationship Map
[Cross-attendee connections — shared employers, mutual connections, overlapping
  interests, organizational dynamics between attendees. Skip if single attendee.]

## Company Context
- **What they do:** [One line]
- **Size / Stage:** [Employees, funding stage, HQ]
- **Recent news:** [Top 2-3 items, dated with source]
- **Relevant to your meeting:** [How their company context connects to your
  discussion — e.g., recent product launch you might discuss, funding that
  signals growth, leadership change affecting priorities]

## Value Positioning
*[Only for sales/discovery, partnership, and board/investor meetings. Omit entirely
  for interview, internal, and general external meetings.]*

### Value Mapping
[Match their specific needs/pain points to your capabilities. Every mapping must
  be grounded in research from Step 4.5, not generic claims.
  Format: "They [specific finding with source] → Your product [specific capability]"]

### Integration Hooks
[Tools/platforms they use that your product integrates with. Only include
  integrations confirmed from research (their tech stack) AND your profile
  (integration_partners). If no overlap found, say so honestly.]

### Recommended Positioning
[2-3 sentences on how to frame your pitch for THIS specific company and person.
  Consider: their company stage, recent news, the attendee's role and priorities,
  and any competitive displacement opportunity. This is the "elevator pitch
  calibrated to this meeting" paragraph.]

### Reference Customers
[Similar companies from your case_studies that match their industry, size, or
  use case. Include the outcome/metric if available. If no matching case studies,
  omit this subsection rather than forcing a weak match.]

## Talking Points
[3-5 specific, actionable conversation starters grounded in the research.
  Not generic "ask about their priorities" — specific: "Ask about their
  migration from [old tool] to [new tool] that they announced last month."
  When value positioning is active, weave 1-2 positioning angles into the
  talking points naturally — don't make every talking point a sales pitch.]

## Watch Out For
[1-3 things to be aware of — sensitive topics (recent layoffs, bad press),
  potential awkward overlaps, information gaps you couldn't fill.]

## Sources
[Numbered list of key URLs cited in the briefing]

Meeting type adaptations:

Core rules:

• Every factual claim about an external company or person must have a source URL. Data drawn from the user's own business profile (differentiators, integrations, case studies) should be attributed to the profile rather than requiring an external source.
• Lead with the Quick Take — most readers stop there.
• Talking points must be specific to THIS meeting, grounded in research findings. Never generate generic conversation starters.
• Say "no public information found" for a person rather than speculating about their role or background.
• If memory has prior meeting notes, surface open items and continuity points prominently — this is the highest-value content.
• Value Positioning claims must be grounded in research from Step 4.5. Never generate generic positioning advice like "highlight your product's strengths." Every value mapping must reference a specific finding about the attendee's company paired with a specific capability from the user's profile or research.
• If sales_context is missing from the profile, note it once at the end of the Value Positioning section: "Tip: Edit your profile at ~/.nimble/business-profile.json to add sales context (differentiators, integrations, case studies) for richer positioning next time."

Step 7: Save to Memory

Make all Write calls simultaneously:

• Report → ~/.nimble/memory/reports/meeting-prep-[company-slug]-[date].md
• Per attendee → ~/.nimble/memory/people/[name-slug].md (use the format in references/memory-and-distribution.md). Add [[path/entity]] cross-references for the attendee's employer (e.g., [[competitors/widgetco]] or [[companies/widgetco]]) and any other discovered relationships.
• Company profile → update ~/.nimble/memory/companies/[company-slug].md if new company data was found. Add reverse cross-references to the people researched (e.g., [[people/alex-kim]]).
• Profile → update last_runs.meeting-prep in ~/.nimble/business-profile.json (only if profile exists)
• Follow the wiki update pattern from references/memory-and-distribution.md: update index.md rows for all affected entity files, append a log.md entry for this run.

The person profile in people/ should contain structured key facts (role, background, interests, communication style) that can be loaded by future meeting prep runs.

Step 8: Share & Distribute

Always offer distribution — do not skip this step. Follow references/memory-and-distribution.md for connector detection, sharing flow, and source links enforcement.

Step 9: Follow-ups

• Go deeper on an attendee → more focused person research
• Add attendees → research additional people joining the meeting
• "What about [topic]?" → targeted search on specific dimension
• "Looks good" → done
• Sibling skills: company-deep-dive for a full 360 on the company, competitor-intel to track them as a competitor, competitor-positioning to compare messaging before a sales meeting

Agent Teams Mode (Dual-Mode)

Check at startup: echo $CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS

Team mode (flag set): Spawn teammates instead of sub-agents. Each teammate researches one attendee and can message the others when finding cross-connections.

How cross-attendee discovery works:

1. Each teammate researches their assigned attendee independently
2. When a teammate discovers a workplace, school, or connection that overlaps with another attendee, they send a message to that teammate: "My attendee [Name] worked at [Company] from 2019-2022 — did yours overlap?"
3. The receiving teammate checks and responds
4. Lead (you) collects all cross-references and builds the Relationship Map section

This produces higher-quality relationship maps than solo mode because teammates actively search for connections rather than just comparing results post-hoc.

Solo mode (flag not set): Standard sub-agent flow from Step 3.

What This Skill Is NOT

• Not competitor monitoring — use competitor-intel for tracking competitors
• Not a company deep dive — use company-deep-dive for research without attendees
• Not a CRM — gathers web intelligence, doesn't manage contacts or pipelines
• Not a calendar app — reads events for context but doesn't manage them

Error Handling

See references/nimble-playbook.md for the standard error table. Skill-specific errors:

• Person not found: Try name variations (full, first+last, with company). If still nothing: "Couldn't find public info on [Name]. Can you share their title or LinkedIn?"
• Ambiguous name: Present top candidates with company/title context and ask.
• Empty company results: Note it and focus on attendee-level findings.