Partner Relationship Intelligence¶

Capability Slug: partner_intelligence Plan Tier: Command Credit Cost: 10 / 12 / 15 credits per run (varies by intelligence type) Category: Co-Sell Default mode: relationship_summary

Overview¶

You already have data about your partners — meetings, emails, deal stages, co-sell registrations, engagement signals. What you don't have is time to read all of it and surface what matters. Partner Relationship Intelligence is the AI synthesis layer that turns that data into structured outputs: relationship summaries, opportunity lists, sales briefs, intro path suggestions, and engagement-health reads.

The capability runs on the data you provide as input. It is not a CRM, not a LinkedIn integration, and not a connection-graph traversal — it's Claude reasoning over the partner data and brand context you pass in.

What the warm-intro-paths mode does (and doesn't)

The warm_intro_paths mode synthesizes intro suggestions from the partner data, interaction history, and intermediary signals you pass in. It does not access LinkedIn connection graphs, anyone's contact list, or external relationship databases. Treat it as "given what I've told the model about my network and my target, what intro paths look most promising?" — useful for structured thinking, not magic relationship discovery. For LinkedIn-specific intelligence, see LinkedIn Ad Intelligence, which uses the live Ad Library API.

The five intelligence types¶

Type	Credits	What you get back	When to use
`relationship_summary` (default)	10	Executive summary, relationship score 1-10, highlights, red flags, next actions	Weekly check-in prep
`co_sell_opportunities`	15	Prioritized opportunities with deal size estimates, partner role, AWS co-sell potential, ACE/MDF eligibility, Private Offer fit	Quarterly co-sell pipeline review
`sales_brief`	10	Pre-meeting brief with elevator pitch, value prop, talking points, objection handling, AWS Marketplace context, action checklist	Pre-meeting prep
`warm_intro_paths`	12	Top 3 paths to each target contact with intro templates, timing recommendations, and recommended approach	Strategic outreach planning
`partner_engagement`	10	Health score 1-10, engagement trend (increasing / stable / declining), red flags, green flags, recommendations	Quarterly partnership health review

All modes use Claude 3 Sonnet for the synthesis. Output is structured JSON.

What `relationship_summary` returns¶

{
  "success": true,
  "intelligence_type": "relationship_summary",
  "credits_used": 10,
  "summary": "...",
  "relationship_score": 7,
  "highlights": ["...", "..."],
  "red_flags": ["..."],
  "next_actions": ["...", "..."]
}

The summary is what the prompt frames: relationship history, health, co-sell activity, strategic alignment, and AWS-specific signals (ACE participation, partner referrals, integration depth).

What `co_sell_opportunities` returns¶

{
  "opportunities": [
    {
      "account": "...",
      "stage": "early|qualified|active",
      "estimated_deal_size": "...",
      "partner_role": "referral|reseller|integrator",
      "aws_co_sell_potential": "high|medium|low",
      "recommended_actions": ["..."]
    }
  ],
  "priority_order": [...],
  "aws_co_sell_potential": {
    "ace_eligibility": "...",
    "funding_opportunities": "...",
    "private_offer_potential": "..."
  }
}

The AWS-specific block captures the signals that matter for the AWS partner motion: ACE eligibility, MDF and co-sell credit fit, Private Offer potential.

What `sales_brief` returns¶

A briefing document organized for pre-meeting prep:

Executive summary (top of brief)
Partner overview, current opportunity, key contacts
Partner priorities, talking points, objection handling
AWS Marketplace context (listing advantages, co-sell benefits, Private Offer guidance, CPPO eligibility)
30-second elevator pitch
Pre-meeting prep checklist

What `warm_intro_paths` returns¶

{
  "warm_paths": [
    {
      "target": "...",
      "paths": [
        {
          "intermediary": "...",
          "relationship_strength": "...",
          "best_approach": "email|linkedin|phone",
          "suggested_message": "...",
          "timing": "..."
        }
      ]
    }
  ],
  "intro_templates": {...},
  "recommended_approach": [...]
}

Up to top-3 paths per target, with intermediary relationship-strength assessment, suggested approach, and templated messaging. Again — this is synthesis over the data you provide, not graph traversal.

What `partner_engagement` returns¶

{
  "health_score": 7,
  "engagement_trend": "increasing|stable|declining",
  "red_flags": ["Ghosting", "One-sided relationship"],
  "green_flags": ["Proactive outreach", "Co-marketing"],
  "recommendations": ["..."]
}

The red/green flag heuristics are baked into the prompt — red_flags looks for ghosting, one-sided relationships, lack of deal progression, competitive threats. green_flags looks for proactive partner outreach, customer introductions, co-marketing collaboration, executive engagement.

Required and optional parameters¶

Parameter	Required	Description
`intelligence_type`	Yes	One of `relationship_summary`, `co_sell_opportunities`, `sales_brief`, `warm_intro_paths`, `partner_engagement`
`partner_name`	Recommended	The partner you're analyzing — flows into the prompt
`data`	Recommended	Free-form or structured partner data (interaction history, CRM export, deal stages). String or JSON-serializable array both work.

Brand-voice context (company.name, company.industry, target_audience) is pulled from your company profile automatically.

When to use which mode¶

Question	Recommended mode
"Give me a one-page read on this partner"	`relationship_summary`
"What deals could we run with this partner this quarter?"	`co_sell_opportunities`
"I have a meeting in an hour, brief me"	`sales_brief`
"I need to reach this exec — what's my best path?"	`warm_intro_paths`
"Is our partnership healthy?"	`partner_engagement`

It is not:

A CRM. Pass your CRM data in as the data parameter; the capability doesn't read your CRM directly.
A LinkedIn integration. Use LinkedIn Ad Intelligence for live LinkedIn data.
A persistent partner-health metric system. The 1-10 scores returned by relationship_summary and partner_engagement are per-run AI assessments, not stored KPIs.
A partner discovery tool. Use Co-Sell Partner Discovery & Matching (cosell_partner_matching) to find partners; this capability is for partners you already work with.

API access¶

Exposed through the agent runtime as partner_intelligence. Partner API keys scoped with view_partners and manage_cosell_plans plus the agent capability slug can run all five intelligence modes. See Partner API Capabilities.

Partner Relationship Intelligence¶

Overview¶

The five intelligence types¶

What relationship_summary returns¶

What co_sell_opportunities returns¶

What sales_brief returns¶

What warm_intro_paths returns¶

What partner_engagement returns¶