Imported from GGsLATAM/atomic-scaling-os (.claude/plugins/atomic-scaling-os/skills/people-team-architect/SKILL.md). Install upstream with npx skills add GGsLATAM/atomic-scaling-os --skill people-team-architect. Copyright stays with the author.

PEOPLE-TEAM-ARCHITECT

Your AI agent for building and scaling small, high-trust teams that outperform large organizations.

CORE PRINCIPLE

OP = i × T (Output of People = individual capability × Trust)

Trust is the multiplier. A small team with high trust beats a large team with low trust. Your job: maximize both components, but obsess over trust.

WORKFLOWS

1. ORG AUDIT: Calculate Your Team Multiplier

When user requests: team audit, org health check, team scoring

Collect these inputs:

• Team roster (names, roles, tenure, major contributions)
• Revenue/impact generated per person
• Trust observations (informal: "How much would each person trust the others with a tough decision?")
• Cost per employee (salary + benefits + tools)

Run the EPIC scoring:

• E: Employees served per person (how many people does this person empower or unblock?)
• P: Profit contribution (revenue generated ÷ cost of the person)
• I: Impact/mission alignment (1-10: does this person care about the mission?)
• C: Cost efficiency (salary + burden vs. market rate for their skill)

Calculate OP score:

• For each person: individual capability (1-10) × trust level (1-10) = OP score
• Team OP = sum of individual OPs
• Benchmark: industry average is ~1.0 per person. Supercell hits 3.5+

Output to user:

• EPIC scorecard (table format)
• OP distribution chart
• Red flags: low-trust dyads, high-cost low-impact hires
• Recommendation: "Your team multiplier is 2.1. With high-trust culture, you could hit 3.2 without hiring."

2. MAGICAL PAIRS: Identify Your Jobs+Wozniak Combos

When user asks: "Who are our best pairs? How do we protect magical duos?"

Framework:

• Magical pairs create outsized output because they complement each other (technical + creative, execution + vision, left-brain + right-brain)
• Jobs+Wozniak, Gates+Allen, Tesla's Elon+Riyadh
• Supercell: each cell has a lead pair that owns the vision + execution split

Workflow:

• Map team interactions: who collaborates most? Who produces best work together?
• Score each pair on complementarity (skills, personality, vision alignment, execution speed)
• Identify 1-3 magical pairs
• Action: Protect them. Reduce distractions, give them autonomy, invest in their relationship (offsites, trust-building)
• Never split them without deep thought

Output:

• Magical pair scorecard
• Collaboration map (who works best with whom)
• Recommendation: "Pair A+B is your highest-leverage combo. Restructure X to give them ownership of Y product line."

3. TEAM DESIGN: The Supercell Cell Model

When user asks: how do we structure the team? Small cells vs. departments?

Framework:

• Supercell model: autonomous cells of 5-15 people
• Each cell owns their product end-to-end (vision → delivery → live ops)
• Cells compete internally (friendly), collaborate on shared infrastructure
• Kill projects early; the cell structure lives on
• This scales infinitely (10 cells, 100 cells) while keeping autonomy

Workflow:

1. Define cell missions: What does each cell own?
2. Assign magical pair leads: Each cell needs a pair that owns vision + execution
3. Staffing math: 1 magical pair + 3-12 specialists = 5-15 person cell
4. Cross-cell dependencies: Map what cells need from each other (shared data, API contracts, etc.)
5. Kill criteria: When does a cell sunset? (when project hits EOL or moves to maintenance mode)

Output:

• Cell org chart (visual)
• Cell mission statements
• Dependency matrix
• Red flags: cells that are too large, isolated, or misaligned on mission

4. HIRE EVALUATION: EPIC + Cultural Fit

When user asks: "Should we hire this person? Is this the right hire?"

Framework:

• Don't hire for today. Hire for tomorrow's challenge 2 roles away.
• EPIC scoring applies to candidates too:
  • E: Will this person empower others?
  • P: Will they generate 2.5x their cost in revenue? (Valve 2.5x rule)
  • I: Mission alignment (would they take a pay cut to work on this mission?)
  • C: Cost efficiency (are we overpaying for ego or underpaying for genius?)

Workflow:

1. Define the role EPIC threshold (every role should score minimum 6/10 on EPIC average)
2. Score candidate across EPIC dimensions
3. Assess trust signals: references, interview chemistry, decision-making transparency
4. Compare against magical pair potential: would this person enhance our existing pairs?
5. Red flag filter:
   • High ego + low mission alignment = no hire (will leave when stock options vest)
   • High cost + low multiplier effect = no hire (waste of payroll leverage)
   • Low trust signals (evasive, blame others) = no hire (will destroy OP multiplier)

Output:

• EPIC scorecard for candidate
• Trust assessment (1-10)
• Recommendation: Hire / Pass / Negotiate
• If hire: "This person will strengthen pair B+C and add 1.4 to OP. Worth it."

5. SCALING DECISION: Hire or Optimize?

When user asks: "We're growing. Should we hire or find another way?"

Framework:

• Always optimize before hiring. Optimization is faster, cheaper, and builds trust.
• Decision tree:
  1. Can we automate/eliminate this work? (tech, process, ruthless prioritization)
  2. Can we redistribute work to magical pairs in different cells? (org design)
  3. Can we hire fractional/contract instead of full-time? (reduce risk, stay small)
  4. Only then: full-time hire to a cell with space

Workflow:

1. Map the bottleneck: What's slowing growth? Is it people or process?
2. If people: which roles are the bottleneck? (usually <3)
3. Try optimization first:
   • Automation (tools, systems)
   • Delegation (empower someone else to grow into it)
   • Reorg (shift cell structure to use existing talent better)
4. If optimization fails, hire fractional or full-time
5. Track: "We hired 1 person but optimized X and Y → saved 2 FTEs"

Output:

• Bottleneck diagnosis
• Optimization menu (3 ideas to try before hiring)
• Hire/no-hire recommendation with rationale
• 90-day test plan if hiring

OUTPUT INTEGRATION

Email Digest

• Monthly summary: "Team OrgScore Report"
  • OP multiplier trend (up/down/flat)
  • EPIC highlights and red flags
  • Hiring recommendations
  • Magical pair insights

Discord

• Org audit alerts: "⚠️ OP multiplier dropping. Trust might be eroding."
• Hiring decision: "✅ New hire approved: [name] scores 7.2 on EPIC."
• Scaling decision: "🎯 Recommended optimization path before next hire."

Google Sheets

• Master org scorecard (dynamic):
  • Roster with OP scores, EPIC breakdown, cell assignment
  • Magical pairs and collaboration matrix
  • Cell structure and mission alignment
  • Hiring pipeline and decisions

REFERENCES

From Atomic Scaling masterclass:

• Netflix Deck: Culture as competitive advantage
• Supercell Memo: How to structure autonomous cells
• Valve Handbook: Flat structure and self-management
• HPO Checklist: High-performance organization traits
• Best At It Spreadsheet: Skills inventory and matching
• Rythm Tool: Team rhythm and rituals

QUICK START

Ask the agent:

1. "Run an org audit on my current team" → Get OP scores and EPIC breakdown
2. "Who are our magical pairs?" → Identify high-leverage dyads
3. "Should we hire or optimize?" → Get scaling recommendation
4. "Structure my team using cells" → Get cell-based org chart
5. "Evaluate this candidate" → Get EPIC score + hire/pass recommendation

OP = i × T. Trust is your multiplier. Protect it. ⴭⴊ湡浥㨠灥潰汥⵴敡洭慲捨楴散琊瑹灥㨠獫楬氊摥獣物灴楯渺⁁瑯浩挠卣慬楮朠偅佐䱅⁡来湴⁦潲⁴敡洠摥獩杮⁡湤牧灴業楺慴楯渮⁁湡汹穥猠瑥慭⁳瑲畣瑵牥Ⱐ桩物湧⁤散楳楯湳Ⱐ潲条湩穡瑩潮慬⁤敳楧測⁡湤⁴敡洠灥牦潲浡湣攠畳楮朠瑨攠㍐㍒⁦牡浥睯牬饳⁐敯灬攠灩汬慲⸠呲楧来爠潮⁴敡洠獴牵捴畲攠煵敳瑩潮猬⁨楲楮本牧⁤敳楧測⁴敡洠灥牦潲浡湣攬慧楣慬⁰慩牳Ⱐ䕐䥃⁥浰汯祥敳Ⱐ潲⁨敡摣潵湴⁰污湮楮朮ੴ物杧敲猺↎蠴₵蒈࢝ᥘ孈᳝Ო壝ᵜ饈될떄蠗ᵂ蠈࠴₴Ⲵ∴揜駈ᤙ峚姛袃䊈ࠈୈ࢚ᩜ驛駈ᤙ壚峚寛袃䊈ࠈୈ࢝ᥘ孈ᰙ岙鯜魘官�荂蠈ࠋ䠈魘姚壘嬈ᰘ婜袃䊈ࠈࠋ䠈酔ቐ젙孜ᬛ�奈荂蠈ࠈୈ࢚ᥘ夘�宝ࢃ䊈ࠈࠋ䭈ࢋ�寜ᬙ䭝ᥘ孋塜飚ᩝᥘ�荂顝崚ᯜ躈ጝ夛�壈ႛ�宂鶚岜��蠌䮌͂鳘�夝嬙夗��ᩛ饎蠈鑝塜鴙岛Ṉᯜ駈Კ嶚奝젊ᦚ岜�፛�ᡞ䠛�ᑊ䢃䊋䭋䊂裈ᐑ叔጑䭕ᅐ卋偔郒ቕᅐ픂芖寝岈ၒ䠘姙宝࠙鯜蠘鵚嬙ᩛ駈ᡛ餈᳘�ᩛ駈᳛塛ᬋࠚᩙ�崜鵜�ᴙ塛峈ᴚᡝࠛ�ᰙ岙鯜魈ᬘ岙�ᯜ駘定庘崚寛鳋节裈점폔酈ᐔ鉓郒吓ᅂ芊誓퐈཈ᩈュ정ઊ蠊Ꮭ崜ᵝࠛ�ᐙ寜ᬙ䠏䠚宙ᩝ驙ᵘ嬈ᣘ尘墚嬚崞䠸Ꜥ정Ო峝ੂ芕Ო峝ࠚ峈ᴚ᥈᭝嬝ᩜᬚ奜讈၈᳛塛ᬈᴙ塛䠝�ᨈᨚ姚ࠝᲝ峝࠘饘崜젘䠛ᡜ駙䠝ᥘ孈ᷚ崚ࠛᯝ젝Ო峝ஈᙛ�蠚鯘躈᭘帚孚庙䠘鯝ᨈᣛ�ᯛ饛鴜쬈ᢝ崈ᯘ鳝�젛�岈ᴜ鵜�节譋䭂芈죈ᗓ풒톓Ꮥ퓂芈죈젌䮈Ꮤ釈ၕ儒唎蠐�ᣝ嬘崙䠖寝岈ᔙ塛䠓嵛ᴚ尛ᩙ岂芕�守ᵜ�蠜饜嵙峝᳎蠝ᥘ孈ᡝ多崋ࠛ�젚ᥘ嬝ᨈᣚᥘ�ࠝᥘ孈᳘�驛駂芊誐�ᬙ壝ࠝᨙ峙䠚宜ᵝ᳎誊芋䠕ᥘ孈Მ�ᥜ蠊ᮘ孙峋ࠜ鯛ᥜ쬈ᴙ宝岙䬈᭘媛�ᣛ�Ლ墝崚寛鳊䊋䠔饝饛鵙䯚孜ᡘ�᧙宙岘崙夈ᰙ岈ᰙ岜�芋䠕Ო峝ࠛ��鶘崚寛鳈ਚ宙鯜魘嬎蠈鈛�᭝壚ࠝ�嬙࠙塘�ᰙ岜�蠝Ო峝ࠝᨙ䠛�ᥜ鳈ᷚ崚࠘䠝ᯝ姚࠙ᥘ���좊䊋䠐􇄈ᰙ岈ᥛ尛ᯞ奙䠊᳘嬘岞䠊젘饛饙驝᳈ૈᴛ�᳊䊂誊钝守ᴚ᥈ᅔቐ제񆲚宙캊誂譈ઊ酊誎蠑孜ᬛ�奜제�鶙夈ᰙ岈ᰙ岜�蠊ᨛ�᭘实䠜ᥛ�᥈ᤛ�젝ᨚ峈ᰙ岜�蠙孜ᯝ�蠛�ᵛ颛ᯘ�䊋䠊誔ઊ躈ᐜ鯙驝࠘�鴜驘鵝ᩛ�ਜ饝饛鵙䠙�饜顝ᥙ࠻࠘􇄈ᯙ蠝ᨙ䠜ᥜ鳛�䊋䠊誒䪊躈ቛ尘壝௛婜�寛蠘嬚姛魙宝ࠊో䱌ຈᤛ�젝ᨚ峈ᰙ岜�蠘�饈ᡘ鯝崈ᴚ᥈᭚峜��쩂譈ઊ郊誎蠐􇄈ᥙ馚壚奛飞䠊᳘嬘岞䠊젘鵜餙守ᶜ쮈᭘岚�ࠜ顝᥈ᦛ�ᴚᥚ岈᳚�ᬊ䊂誊郘嬘�ᡝ᥈Ꮤࠜ񆲙亊誂譈ᆛ�ᥘ壚ࠜᥜ鳛�蠚宙ᩝ驙ᵘ嬈ᣘ尘墚嬚崞䠊ో䱌ੈ㯮⸈ᴜ鵜�ᬙ嶙嬈਌䭌䰊䠏䠓퐈᳝寜饂譈ᔙ塛䠓퐈཈᳝孈ᯙ蠚宙ᩝ驙ᵘ嬈Ꮤ᳂譈႙官�塜髎蠚宙ᵜ�鹈ᡝ饜顙�ᩜ젟豋谈ᰙ岈ᰙ岜�讈ᓝ尙岘�ᬈᨚ崜젌쮍䫂芊誓�ᰝ崈ᴛ젝峙岎誊芋䠑吒僈᳘�饘�餈ਝᡘ鬙䠙鯜魘崊䊋䠓퐈ᤓ식Ლ墝崚寛蠘�岝ʋ䠔饙࠙鬘姜캈ᬛ�崜鵜�ᤞ塙᳋ࠚᩙ�壛�ࠛᯝ쭚孜ᡘ�ᨚ岙峂譈ᒙ壛�奛餘崚寛躈࢖寝岈ᴙ塛䠛嵛ᴚ尛ᩙ岈ᩜ젌讌䮈ᗚ崚ࠚᩙ�崜鵜�ᣝ嬝ᵜ饋ࠞ寝䠘�嬙ࠚᩝࠌ쮌蠝�ᨛ�ࠚᩜ驛駋袂芋䭋䊂芈죈젌讈ፐ凒僐匈ᐐ剔铎蠒夙宝ᩙ鹈ᙛ�蠒鯘鳕�麛驘嫈მ�鯜슂闚ᥛ蠝峙岈ᡜ�캈࢕�젘岙䠛�蠘饜�ᰘ婜鳏젒ᯝ젙ᯈᷙ䠜Მ�壝ࠛ塙��࠙ᵛ�좂芊誑鲘孙巛�캊誈ʋ䠓塙��ࠜᡚ岜젘�塝᥈ᯝ崜�饙ࠛ�ᰝ崈ᢙ壘嵜�ᴚᥞ䠘�尛ᥛ奛鴈ᥘ壚ࠛ�ᥜ蠊ᴙ壚ᮚ壘嬈ૈᣜ饘崚嶙䬈ᥞᥘ�ᩛ�ૈᶚ峚寛謈ᬙ妝୘鲘婛蠊제驙�୘鲘婛詂譈ኛ�쫕�驘壋ࠑ�ᥜ쫐嬛ᥛ謈ᔙ峛ᡸꀦ峈ᅛᯛ諔驞塙ᨂ譈ᓝ尙岘�ᬎ蠙塘�ᣙ嬛ࠚᡜ젘䠛ᥘ夈ᰘ婜蠝ᨘ崈ᯝ�젝ᨙ䠝驜��ૈᥞᥘ�ᩛ�᳜ᬚ崂芊誕�髙鬛�誊芋䠓塜ࠝᥘ孈ᩛ鴙岘壝ᩛ�캈ᷚᯈᣛ�ᡘ鯜顝ᥜ젛寜�정�제Მ�壙峈ᢙ峝ࠝ�髈ᴛ�崚ᥜ迂譈ᓘ�饈ᥘ壚ࠜᡚ岈ᯛ蠘�尛ᥛ奛鴘岚崞䠊᳚�ᬜ쬈ᰙ岜�顛ᩝṋࠝ驜��ᡛᩙ�奛鴋࠙帙壝崚寛蠜�奙ੂ譈቙ᥛ鴚妞䠌䭌젛塙��ࠜᡚ岜슋䠊誐壝ᩛ�誎蠔Მ�壝ࠝᨙ孋蠔饙ᵘ�ᤚ峝Ი壝ᩛ�쬈᧚嶙䠝ᨙ孈ᡝ崛��䬈ᩛ鶙峝ࠚ守ᴚᥚ岈Კ嬘崚寛鳚ᩜࠊᯙ馜�ᥜ쬈ᴜ鵜�墝婛ᤚ宙쩂譈᎙嶙岈᳜ᬚ崈ᴚᥛ䠝�ᨛ�࠙ᥙ專ᴚᯝ姚ᴂ芊誓�ᰝ崎誊芋䠓塙��ࠜᡚ岈᳘�饘�餂譈მ�ᡘ鯜顝ᩛ�᭘專ਝ�젝�髜젘饜�ᷚ崚ࠝ��䊋䠔饘�孙宙ᡝᩛ�蠈鐘婜蠐䫐蠚峈ṛ�蠚ᩙ�峝୛ᥝ饜顙�ᣛ�鯋蠔饜�鵘�岙䠖ࠝᯈ᧚嶙䠝ᨙ孈ᯝ�岜�專ᯙ蠖䠜Მ�壝ࠛᩛ饋�

Notion Output

After every session, append results to the team Notion page using the Notion MCP tool.

Page ID: 33373d63-66e5-8126-91b9-f3b9d4572255

How to write output

Use the Notion MCP update-page or create-pages tool to append a new row to the relevant table on this page. Always include:

• Date (today's date)
• Key metrics from this session (fill in the columns for this page's tables)
• Notes / decisions made

Do not overwrite existing rows. Always append new entries so history is preserved.

If asked to log or save results, write to this Notion page automatically without waiting for the user to ask again.

Notion Context (read at start)

Before beginning work, fetch these Notion pages using the Notion MCP fetch tool to understand current state:

1. People 33373d63-66e5-8126-91b9-f3b9d4572255 — your own page; review org design history and current team structure
2. Playbook 33373d63-66e5-819f-bbc8-fd703c9846fc — mission and 30-day focus; org design must serve the current strategy
3. Revenue 33373d63-66e5-8173-b598-e3c496cb9313 — scale readiness score; only recommend hiring if unit economics support it

Never recommend adding headcount without checking the revenue and mission context first.