---

name: matchmaker description: > Scan your agent stack, map capability gaps, and recommend Claw Mart products ranked by impact. USE WHEN: user says "analyze my agent setup", "what personas should I get", "what am I missing", "recommend personas for my workflow", "where are the gaps in my agent capabilities", "I have $X to spend", or wants to understand their agent stack coverage. DON'T USE WHEN: user wants to search the Claw Mart catalog directly (use clawmart-manager), wants to install or download a specific skill, or wants security analysis of a skill (use skill-inspector). OUTPUTS: capability coverage map, ranked gap analysis, and top persona recommendations with reasoning, prices, and purchase links.

Matchmaker

Version: 1.1.0 Price: Free Type: Skill

Description

Smart recommendation engine for your OpenClaw agent stack. Matchmaker reads your installed skills and personas, maps them against 19 capability domains, and identifies the gaps that are costing you the most productivity. Then it recommends the exact Claw Mart personas that would fill those gaps — ranked by impact, not by price.

Think of it as a personal shopper for your agent setup. It looks at what you have, understands what you do, and tells you what you're missing.

Prerequisites

• An OpenClaw installation with at least one skill or persona installed
• Local skills directory accessible (default: skills/ in your OpenClaw root)
• Optional: CLAWMART_API_KEY=cm_live_... for live catalog availability checks

Setup

1. Copy SKILL.md into your OpenClaw skills directory (e.g. skills/matchmaker/SKILL.md)
2. Reload OpenClaw
3. Test with: "Analyze my agent setup"
4. Optional: set CLAWMART_API_KEY to enable live catalog checks and direct purchase links

Commands

• "Analyze my agent setup"
• "What personas should I get?"
• "What am I missing in my agent stack?"
• "Recommend personas for my workflow"
• "Where are the gaps in my agent capabilities?"
• "I have $20 to spend — what should I buy?"
• "What would complement my current skills?"
• "Show my capability coverage"

How It Works

Analysis

Matchmaker scans your local skills/ directory and reads every SKILL.md it finds. For each installed skill or persona, it extracts:

• Type — skill vs. persona
• Capabilities — the declared command set and stated scope
• Domain signals — keywords, patterns, and role descriptions that map to capability domains
• Overlap — where multiple installed packages cover the same domain

It builds a capability profile: a snapshot of what your agent can currently do, organized by domain.

Gap Detection

Matchmaker evaluates your profile against 19 capability domains. For each domain, it assigns a coverage level:

| Coverage | Meaning | |----------|---------| | Strong | You have a dedicated persona or multiple skills covering this domain | | Partial | A skill touches this domain but doesn't specialize in it | | None | No installed package addresses this domain |

Gaps are domains at None or Partial coverage. Not every gap matters — Matchmaker weighs gaps by how relevant they are to your actual workflow, based on what your existing tools suggest you do.

Recommendation Engine

Recommendations are scored on three factors:

1. Gap severity (50%) — How uncovered is this domain in your current setup? A domain with zero coverage scores higher than one with partial coverage.
2. Adjacency (30%) — How closely does this persona complement what you already have? A security persona is more valuable if you already have DevOps tools but no security coverage.
3. Cost efficiency (20%) — What's the capability-per-dollar? A $9 persona that fills a critical gap outranks a $14 persona that fills a minor one.

When a budget is specified, Matchmaker optimizes the combination of personas that maximizes total gap coverage within the stated budget.

Capability Domains

| Domain | What It Covers | Persona | |--------|---------------|---------| | Security — Defense | Vulnerability scanning, hardening, code review for exploits | Sentinel ($9) | | Security — Offense | Penetration testing, attack surface analysis, threat modeling | Breach ($14) | | DevOps & Cloud | Infrastructure provisioning, CI/CD, Docker, Terraform, secrets management | Forge ($14) | | SRE & Incident Response | Monitoring, incident triage, postmortems, SLO tracking | Pulse ($9) | | Product Management | PRDs, backlog prioritization, sprint planning, OKRs | Compass ($14) | | Launch & Release | Launch checklists, release sequencing, announcements | Ship It ($14) | | Content & Writing | Blog posts, newsletters, social threads, voice calibration | Ghost Writer ($9) | | Social Media | Platform-native posts, content calendars, engagement threads | Megaphone ($9) | | Sales & Outreach | Prospect lists, outreach sequences, lead scoring, proposals | Deal Flow ($14) | | Market Research | Competitor analysis, positioning maps, pricing intelligence | Recon ($9) | | Finance & Accounting | Transaction categorization, cash flow forecasting, financial reports | Ledger ($14) | | Growth Strategy | Growth experiments, funnel analysis, retention loops | Signal ($14) | | Legal & Compliance | Terms of service, contract review, compliance checklists | Counsel ($9) | | QA & Testing | Test plans, edge case identification, regression suites | Probe ($9) | | Data & Analytics | Database queries, dashboards, trend forecasting, executive reports | Cortex ($14) | | Documentation | API references, tutorials, changelogs, onboarding docs | Blueprint ($9) | | Community & DevRel | Community updates, developer guides, event planning, engagement metrics | Relay ($9) | | Customer Support | Help center articles, support responses, escalation playbooks, CSAT tracking | Closer ($9) | | Agent Building | Designing and writing production-grade personas and skills for OpenClaw | Architect ($19) |

Output Format

=== MATCHMAKER REPORT ===
Generated:  [timestamp]
Skills scanned: [N] skills, [N] personas

CURRENT COVERAGE
  Security — Defense    ████████████████████ Strong   (Sentinel)
  DevOps & Cloud        ████████████░░░░░░░░ Partial  (custom deploy scripts)
  Product Management    ░░░░░░░░░░░░░░░░░░░░ None
  QA & Testing          ░░░░░░░░░░░░░░░░░░░░ None
  Content & Writing     ████████████████████ Strong   (Ghost Writer)
  ...

GAPS IDENTIFIED (ranked by impact)
  1. DevOps & Cloud — You have deployment scripts but no infrastructure management.
  2. Product Management — No PRD or backlog tooling. You're building without a map.
  3. QA & Testing — No test planning or regression tracking.

TOP 3 RECOMMENDATIONS
  #1  Forge ($14) — DevOps & Cloud Engineer
      Why: Your deploy scripts handle the last mile, but Forge handles everything
      upstream — Dockerfiles, Terraform, CI/CD pipelines, secrets rotation.
      You're doing half the job manually.
      https://www.shopclawmart.com/listings/forge

  #2  Compass ($14) — Product Manager
      Why: You have engineering and content covered but nothing driving
      prioritization. Compass writes PRDs, manages backlogs, and defines
      the metrics that tell you what to build next.
      https://www.shopclawmart.com/listings/compass

  #3  Probe ($9) — QA & Test Engineer
      Why: You're shipping without a safety net. Probe builds test plans,
      catches edge cases, and tracks regressions across releases.
      https://www.shopclawmart.com/listings/probe

TOTAL: $37 | Budget: not specified

Budget-Aware Output

When a budget is specified, the report adapts:

BUDGET: $20

OPTIMIZED PICKS (maximizing coverage within budget)
  #1  Forge ($14) — fills your highest-severity gap (DevOps & Cloud)
  #2  Probe ($9)  — fills QA & Testing for $9, staying within budget

  Subtotal: $23 — over budget by $3.
  Adjusted: Drop Probe, add Sentinel ($9) for Security — Offense partial.
  Final: Forge ($14) + Sentinel ($9) = $23 — still over.
  Best fit: Probe ($9) + Sentinel ($9) = $18 — two gaps closed, $2 remaining.

  OR: Forge ($14) alone covers your #1 gap with $6 to spare.

  Matchmaker recommends: Forge ($14) — single highest-impact pick within budget.

Recommendation Logic

| You Have | You're Missing | Recommend | Because | |----------|---------------|-----------|---------| | Sentinel (defense) | Offense testing | Breach ($14) | Defense without offense is half the picture — you can harden but can't verify | | Ghost Writer (content) | Distribution | Megaphone ($9) | Great content with no distribution channel sits unread | | Deal Flow (sales) | Market context | Recon ($9) | Outreach without market intelligence means shooting blind | | Forge (DevOps) | Incident response | Pulse ($9) | You can build infrastructure but not respond when it breaks | | Compass (product) | Launch execution | Ship It ($14) | PRDs are worthless if nothing ships — Ship It closes the gap | | Ship It (launch) | Product direction | Compass ($14) | You're shipping fast but may be shipping the wrong things | | Signal (growth) | Sales execution | Deal Flow ($14) | Growth strategy without pipeline execution leaves revenue on the table | | Cortex (data) | Growth strategy | Signal ($14) | You have the numbers but no framework for acting on them | | Probe (QA) | Documentation | Blueprint ($9) | Tests prove it works, docs prove someone can use it | | Blueprint (docs) | QA coverage | Probe ($9) | Documentation without testing means documenting broken behavior | | Relay (community) | Support ops | Closer ($9) | Community engagement without support infrastructure burns goodwill | | Closer (support) | Community building | Relay ($9) | Support handles problems, community prevents them | | Any technical stack | No legal review | Counsel ($9) | Shipping without compliance review is a liability waiting to happen | | Any go-to-market stack | No financial ops | Ledger ($14) | Revenue without financial tracking is growth without visibility | | Nothing installed | Starting fresh | Architect ($19) | Build the agent stack itself — Architect designs the skills and personas you need |

Guardrails

• Honest recommendations only. If the user's setup has no meaningful gaps, say so. Never recommend a persona just to recommend something.
• Never pushy. Present findings and reasoning. The user decides. No "you need this" or "don't miss out" language.
• Always show reasoning. Every recommendation includes a specific explanation tied to the user's actual setup — not generic marketing copy.
• Respect budget constraints. When a budget is stated, optimize within it. Never recommend over budget without explicitly flagging the overage and offering alternatives.
• No duplicate recommendations. Never recommend a persona the user already has installed. Check the skills directory first.
• Transparent scoring. If asked, show the raw gap severity, adjacency, and cost efficiency scores that produced the ranking.
• Catalog accuracy. Only recommend personas that exist in the current Claw Mart catalog. Never invent or speculate about personas that don't exist.

Why Free?

Helping you find the right tools for your workflow benefits the whole ecosystem. Better-matched agent stacks mean happier users, better reviews, and personas that get used instead of shelved.