agentsclimarketplace

Product design

Skill coco-research/coco/systems/superintelligence/product-design

Meet Coco. A superintelligent agent framework powered by an advisory board of 389 world-class minds. Scale your AI assistant into a complete engineering department with 142 skills, 277 commands, and persistent state. Universal compatibility. Local privacy. Free and open source.

Install
npx -y skills add coco-research/coco --skill product-design

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.

SKILL.md

10.0 KB, ~2.8k tokens by cl100k_base, as published. Nobody here has run it

Product & Design Super Intelligence Team — SKILL entry

Rijul's product + design brain trust and decision-making partner. 56 named native personas across 8 cells — the people who defined product strategy, discovery, growth, usability, design craft, content, design systems, and the behavior/ethics frontier — plus Adam Wathan cross-listed from Engineering. One merged team (product and design held in deliberate tension) invoked by the /SI-PD-* slash commands. Primary purpose: help take product and design decisions, not just review afterward.

Status: Roster build complete (2026-06-01).

  • Merge decision: Product and Design were combined into ONE team (mirrors the Cloud+Code → Engineering merge). Discipline-pure cells preserve depth; the team convenes both lenses by default. Pure-discipline input stays available via /SI-PD-Huddle <cell> or --cells.
  • Roster: 56 native across 8 cells, locked after a /ultra-think pass on the draft 50 (added Ben Thompson, Ronny Kohavi, Tristan Harris, BJ Fogg, Léonie Watson, Clayton Christensen, Eric Ries, Paula Scher, Susan Kare; swapped out van Schneider).
  • Commands: 25 /SI-PD-* installed at ~/.claude/commands/, generated by python3 superintelligence/ai/scripts/build_commands.py --team PD.

Machine source of truth: registry.json, regenerated from persona frontmatter by python3 superintelligence/product-design/scripts/build_registry.py.

What this team is for

When a CoCo prompt is a product or design decision — what to build, whether the flow is right, what to cut, how to grow, whether an engagement mechanic is ethical — this team supplies named external voices with documented, citable stances. Every claim is attributed to a specific person. Built-in tensions the orchestrator can surface:

  • Engagement vs ethics: Nir Eyal (Hooked) ↔ Tristan Harris (Center for Humane Technology), with BJ Fogg as the academic root.
  • Three JTBD camps: Teresa Torres (continuous discovery) ↔ Tony Ulwick (rigorous ODI) ↔ Bob Moesta (switch interviews).
  • Discovery vs growth: Marty Cagan / Melissa Perri ↔ Elena Verna / Andrew Chen.
  • Quant vs qual: Ronny Kohavi (experimentation) ↔ Jakob Nielsen / Don Norman (usability, humane design).
  • Minimalism vs maximalism: Dieter Rams / Jony Ive ↔ Paula Scher.
  • AI-will-fix-a11y vs accessibility-as-craft: Jakob Nielsen ↔ Léonie Watson.

Cells (8)

CellCountFocusFile
product-strategy7Product vision, strategy, positioning, business modelcells/product-strategy.md
product-discovery-research8Continuous discovery, JTBD, user research, leancells/product-discovery-research.md
growth-metrics8Growth loops, metrics, experimentation, PLG, analyticscells/growth-metrics.md
design-foundations-usability7HCI, usability, interaction design, accessibilitycells/design-foundations-usability.md
design-leadership-craft8Design leadership, visual/industrial craft, culturecells/design-leadership-craft.md
ux-content-research5Content strategy, content design, information architecturecells/ux-content-research.md
design-systems-interaction5 (+1 cross-listed)Design systems, component architecture, patternscells/design-systems-interaction.md
sprints-behavior-bridge8Product×design: sprints, behavioral design, ethicscells/sprints-behavior-bridge.md

Personas (56 native)

Each has a YAML-frontmatter profile under personas/<slug>.md plus a research dump under research/<slug>/. (archetype) = persistent_signals rather than recent signals.

Product Strategy (7): marty-cagan · shreyas-doshi · melissa-perri · april-dunford · gibson-biddle · des-traynor · ben-thompson

Product Discovery & Research (8): teresa-torres · tony-ulwick · bob-moesta · indi-young · steve-portigal · erika-hall · clayton-christensen (archetype) · eric-ries

Growth & Metrics (8): lenny-rachitsky · elena-verna · andrew-chen · casey-winters · brian-balfour · sean-ellis · crystal-widjaja · ronny-kohavi

Design Foundations & Usability (7): don-norman · jakob-nielsen · alan-cooper (archetype) · kim-goodwin · steve-krug (archetype) · bruce-tognazzini (archetype) · leonie-watson

Design Leadership & Craft (8): julie-zhuo · john-maeda · jony-ive · dieter-rams (archetype) · michael-bierut · aarron-walter · paula-scher · susan-kare

UX, Content & Research (5): kristina-halvorson · sara-wachter-boettcher · karen-mcgrane · jared-spool · abby-covert

Design Systems & Interaction (5 native + 1 cross-listed): brad-frost · nathan-curtis · alla-kholmatova (archetype) · josh-clark · dan-mall · adam-wathan (cross-listed from Engineering)

Sprints, Behavior & Bridge (8): jake-knapp · john-zeratsky · luke-wroblewski · braden-kowitz · nir-eyal · tristan-harris · bj-fogg · kathy-sierra (archetype)

Cross-listed (1)

adam-wathanteams: [engineering-super-intelligence, product-design-super-intelligence], home_team: engineering-super-intelligence. File lives under superintelligence/engineering/personas/; referenced here via the registry's cross_listed field. No duplication.

Slash commands (25 — installed)

Orchestrator-first, identical surface to the AI and Engineering teams. Generated by build_commands.py --team PD.

  • Dispatcher / Orchestrator: /SI-PD · /SI-PD-Orchestrate "<prompt>" (scores all 56 by domain 40% + cell-coverage 30% + productive-conflict 30%, picks 16–32, approval gate, re-picks every invocation).
  • Identity (skip orchestrator): /SI-PD-Ask <slug> "<q>" · /SI-PD-Huddle <cell> "<topic>" · /SI-PD-Meeting "<prompt>" · /SI-PD-Read <slug>.
  • Roster: /SI-PD-Recruit <domain> "<why>" (reads ROSTER.md for known gaps).
  • Action (15): Analyse · Decide (primary) · Review · Re-Analyse · Pre-Mortem · Post-Mortem · Full-Cycle · Tradeoff · Plan · Design · Vote · Debug · Stress-Test · Defend · Roast.
  • Maintenance: /SI-PD-Refresh · /SI-PD-Verify · /SI-PD-VoiceCheck.
  • Global flags: --no-orchestrate · --cells <list> · --personas <list>.

Files in this team

superintelligence/product-design/
├── SKILL.md                   This file.
├── ROSTER.md                  Locked roster + build-wave manifest.
├── registry.json              Machine source of truth.
├── personas/                  56 *.md native personas.
├── cells/                     8 *.md cell summaries (build_cells.py).
├── research/                  56 research dumps.
└── scripts/
    ├── build_registry.py
    └── build_cells.py

Templates (persona.md, convene.md) shared at superintelligence/templates/.

Conventions

  • Schema source of truth: superintelligence/templates/persona.md.
  • Regenerate: build_registry.py then build_cells.py after any persona edit.
  • Attribution at every line; every public_stance carries an evidence_url.
  • Caveman mode never applies to persona/cell/SKILL files — full English prose.
  • Cross-team personas: single file, teams: [...] array + home_team pointer; never duplicated.

Build provenance

  • Built 2026-06-01 by Rijul Kalra during a /coco session, after AI and Engineering.
  • Merge (Product + Design → one team) and the roster patch decided via /ultra-think.
  • 56 native personas built by parallel research sub-agents across 4 waves; agents corrected many stale affiliations against verified 2026 facts (e.g., Elena Verna now Head of Growth at Lovable; Casey Winters now CEO of SuperMe; Karen McGrane at Contentful; John Maeda CVP Design & AI at Microsoft; Nathan Curtis at Directed Edges; Jony Ive's OpenAI device delayed to 2027 and renamed off "io"; Susan Kare built active, not archetype, on found recent signals).
  • Quality bar matched the Karpathy reference: ≥8 cited URLs, ≥3 recent signals (or persistent_signals for archetypes), every public_stance cited, full narrative.

What ships with it: 124 files

2094.6 KB alongside SKILL.md, 2 of them executable

84 more files not listed here. See all 124 in the repository.

Keep looking

Skills are one crate of 327,069. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.