Agent skill

Team Composition

by manager-dot-dev in manager-dot-dev/manager-skills

Helps engineering managers diagnose team skill gaps and make better hiring and assignment decisions — produces the Dungeon Party archetype model (Warrior, Tank, Healer, Wizard, Rogue), the Barrels…

MITAuto-check passed

Install Team Composition

skills CLI
$ npx skills add manager-dot-dev/manager-skills --skill team-composition -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install manager-dot-dev/manager-skills team-composition --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/manager-dot-dev/manager-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/team-composition .claude/skills/team-composition && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
team-composition
GitHub stars
114
Token cost
~2.5k tokens
SKILL.md length
1,393 words
Files
2 (incl. references)
Skills in repo
27
Repo updated
First seen
Licence
MIT

At a glance

Helps engineering managers diagnose team skill gaps and make better hiring and assignment decisions — produces the Dungeon Party archetype model (Warrior, Tank, Healer, Wizard, Rogue), the Barrels…

  • Works in 5 steps: Team need: execution, reliability, glue,… → Current imbalance: which archetype or… → Evidence: symptoms in delivery,… → …
  • The user says team balance
  • SKILL.md covers Before Starting, Response Style, How to Use This Skill and Default Response Shape, plus 8 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Team Composition is an agent skill from manager-dot-dev/manager-skills. Helps engineering managers diagnose team skill gaps and make better hiring and assignment decisions — produces the Dungeon Party archetype model (Warrior, Tank, Healer, Wizard, Rogue), the Barrels and Ammunition framework for understanding throughput limits, the Commandos/Infantry/Police phase model, and a minimum team size guideline. Use when the user says "team balance," "what roles do I need," "who should I hire next," "team is missing something," "skill gaps," "team is slow despite headcount," "this person…

Its SKILL.md is about 2.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including reference files (for example `references/sources.md`).

The repository describes itself as: Skills for engineering managers. The licence is MIT.

When your agent uses it

  • The user says team balance
  • What roles do I need
  • Who should I hire next
  • Team is missing something

Example prompts

  • “team balance,”
  • “what roles do I need,”
  • “who should I hire next,”
  • “/team-composition”

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

  1. Team need: execution, reliability, glue, architecture, exploration, or phase-fit.
  2. Current imbalance: which archetype or capability is missing or overloaded.
  3. Evidence: symptoms in delivery, incidents, planning, collaboration, or morale.
  4. Intervention: hire, grow, reassign, pair, split ownership, or change expectations.
  5. Risk: what happens if the team adds the wrong kind of person.

What it can do on your machine

Read from SKILL.md and the folder at commit c47ebc7. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Team Composition loads about 2.5k tokens when it runs, and up to ~2.8k if it reads all its reference files. Until then it costs about 152 tokens; SKILL.md has 1,393 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~152
When it runs · the whole SKILL.md, loaded when a task matches
~2.5k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~2.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from manager-dot-dev/manager-skills at commit c47ebc7, republished under its MIT licence (© manager-dot-dev). 1,393 words, ~2,549 tokens.

Download SKILL.mdSave it as .claude/skills/team-composition/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
team-composition
description
Helps engineering managers diagnose team skill gaps and make better hiring and assignment decisions — produces the Dungeon Party archetype model (Warrior, Tank, Healer, Wizard, Rogue), the Barrels and Ammunition framework for understanding throughput limits, the Commandos/Infantry/Police phase model, and a minimum team size guideline. Use when the user says "team balance," "what roles do I need," "who should I hire next," "team is missing something," "skill gaps," "team is slow despite headcount," "this person thrived before but struggles now," or "what type of engineer should I hire."
metadata.version
2.1.2

Team Composition

Before Starting

Check for EM context first. If .agents/em-context.md exists, read it.

If .agents/em-context.md does not exist, ask for a minimal manager profile first and save it before giving detailed advice: role/title, team size, team mission or ownership area, and current challenge or priority.

If a specific person is central to the conversation and .agents/reports/[name].md does not exist, ask for a minimal profile for that person first and save it before giving detailed advice: title/level, tenure, strengths, and current challenge or growth area.

If the conversation reveals durable new context later, update .agents/em-context.md or .agents/reports/[name].md automatically. Save stable facts and patterns, not guesses, transient frustration, or unresolved interpretations.

Response Style

Keep the first answer concise and useful. Do not dump the whole framework unless the user asks for depth.

Default to:

  • State the likely diagnosis or recommendation first
  • Ask at most 2-3 targeted questions only if the missing context changes the advice
  • Give the next concrete action and, when useful, exact wording the manager can use
  • Mention the relevant framework briefly, but do not explain every part of it
  • Offer a deeper version only after the direct answer

How to Use This Skill

  • Diagnosing what the team is missing / who to hire next → The Dungeon Party Model
  • Team is slow despite having enough headcount → Barrels and Ammunition
  • Senior engineer who thrived before is struggling now (or vice versa) → Commandos, Infantry, and Police
  • Team is dangerously small or built around one person → Minimum Team Size

Default Response Shape

When helping with team composition, diagnose capability before recommending headcount:

  1. Team need: execution, reliability, glue, architecture, exploration, or phase-fit.
  2. Current imbalance: which archetype or capability is missing or overloaded.
  3. Evidence: symptoms in delivery, incidents, planning, collaboration, or morale.
  4. Intervention: hire, grow, reassign, pair, split ownership, or change expectations.
  5. Risk: what happens if the team adds the wrong kind of person.

Do not default to "hire a senior." Explain what capability is missing.


The Dungeon Party Model

Every role-playing game requires a balanced party. A team of all warriors dies to magic. A team of all healers can't kill anything. The same logic applies to engineering teams.

The five archetypes:

The Warrior

Senior problem-solver. Owns the hard bugs, the production fires, the gnarly migrations. Decisive and effective under pressure. The person you call when something is genuinely broken.

Team need it fills: execution on difficult, high-stakes work.

The Tank

Reliable executor, typically earlier in career, works closely with a Warrior. Not expected to lead — expected to deliver consistently on well-defined work and absorb some of the Warrior's load.

Team need it fills: sustained output on standard work without burning senior bandwidth.

The Healer

Empathy-driven, business-oriented, the team's social glue. Often the bridge between engineering and product/design. Great at facilitating difficult conversations, unblocking cross-team friction, and keeping team morale intact during hard periods.

Team need it fills: relationships, communication, and organizational navigation.

The Wizard

Senior or staff engineer who handles system design, architecture documents, and systemic thinking. Sees consequences several steps ahead. Slower to produce output but prevents expensive mistakes.

Team need it fills: architectural quality and long-range technical direction.

The Rogue

Versatile full-stack utility player. Context-switches across areas, ships exploratory work, covers gaps. The person who can pick up anything and make progress.

Team need it fills: flexibility and coverage in unpredictable situations.


Using the Model

Audit your current roster. For each engineer, identify their primary archetype. Then look for gaps: which archetype is missing or overloaded?

Common gaps:

  • No Tank → Warriors burn out on routine work
  • No Healer → Team is technically strong but organizationally brittle; cross-functional friction accumulates
  • No Wizard → Architecture decisions get made ad hoc; technical debt compounds

Watch for EM archetype bias. EMs tend to over-hire in their own archetype. Former Warriors hire Warriors and ignore Healers. Former Healers hire for culture fit and end up short on execution. The model counteracts this by making the gap explicit.

Use it for hiring decisions. When a headcount opens, map the current team first. The next hire should fill the most critical gap — not replicate the most common archetype.


Caveats

This is a diagnostic model, not a box to lock people into. Most senior engineers span multiple archetypes. The point is not to label people but to identify what the team is currently missing and what to optimize for next.


Barrels and Ammunition

Keith Rabois's insight: adding headcount doesn't linearly increase velocity, because most hires are ammunition — skilled people who need direction. What limits your team's throughput is the number of barrels.

A barrel is someone who can take an idea from conception all the way to shipping and bring people along in the process. They don't need direction — they create it. Adding a barrel doubles your effective output. Adding ammunition without barrels just creates more coordination overhead.

Barrels are rare and culturally specific. Someone who is a barrel at one company may not be at another — the skill is partly about navigating a specific organization. When you find one, treat them accordingly: give them equity, visibility, and significant scope. Replacing a barrel is nearly impossible.

Practical application: when a team is slow despite headcount, the diagnosis is often "not enough barrels." Before requesting more engineers, ask: how many of the people we already have can end-to-end own something? If the answer is one or two, more ammunition won't fix it.


Show full SKILL.md (498 more words)Show less

Commandos, Infantry, and Police

Robert X. Cringely's model of how companies evolve requires different types of people at different stages:

Commandos — the first wave. Operate at startup speed. Work hard, fast, and cheap. Their job is to establish a beachhead: build the prototype, prove the idea, move before anyone knows they're there. They're not precious about quality — they're precious about speed.

Infantry — the second wave. Scale what the commandos started. They work systematically, build properly, and ship the first real product. Where commandos made 20 decisions a day, infantry makes them stick.

Police — the third wave. Companies that succeed eventually need compliance, process, and stability. Police ensure repeatable quality but are poor at innovation or speed.

The management trap: most people strongly prefer one of the three modes and are poorly suited for the others. A commando in a police phase is a troublemaker. A police officer in a commando phase is a bottleneck. Mismatches explain a large fraction of "why is this person struggling?" situations.

Two practical uses:

  1. When hiring, identify which phase your team/product is in. Hire commandos to build new things; hire infantry to scale them. Police skills are rarely a shortage in engineering.
  2. When a senior engineer who thrived in a startup phase struggles in your scaled organization (or vice versa), phase mismatch is often the explanation — not a sudden decline in skill.

Minimum Team Size

From An Elegant Puzzle (Will Larson): never let a team drop below roughly 4 people — and never create a team around a single person.

A team of 1 or 2 is structurally fragile: one person's vacation, illness, or departure can halt everything. It also means no real code review, no knowledge redundancy, and no ability to handle an incident while also doing normal work.

The failure mode to avoid: creating small "teams" as organizational fictions — one engineer "owns" a product area and is technically a team. This engineer carries all the risk and gets none of the support structure a real team provides. When something goes wrong or when they leave, there's no team to absorb it.

Minimum viable team (Larson's guidance): ~4 engineers. This gives enough coverage for on-call, code review, technical design, and vacation without creating a single point of failure.

When you're below this: either grow the team to minimum viability, merge it into a larger adjacent team, or be explicit that it's a temporary solo/pair arrangement with a transition plan — not a permanent structure.


Dive Deeper

If the user asks where a framework came from, wants to read the original article, or wants more context on any topic in this skill — read references/sources.md.


  • hiring — Use the model to define what you're looking for before writing the job description
  • delegation — Kingdom assignments often follow archetype: give systems ownership to Warriors and Wizards, cross-team coordination to Healers
  • managing-high-performers — Wizards in particular need visible, high-complexity work to stay engaged
  • engineer-motivation — Phase mismatch (Commando in a Police phase) often looks like a motivation problem

© manager-dot-dev, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file (references) in skills/team-composition of manager-dot-dev/manager-skills.

  • SKILL.md
  • references/sources.md

Open the folder on GitHubat commit c47ebc7

Compare with similar skills

Team Composition next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Team Composition compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Team Composition this skillmanager-dot-dev/manager-skills114—~2.5kAutomated safety check: PassMIT
Diagnosegithub/awesome-copilot40k1 repos~1kAutomated safety check: PassMIT
Gapsanthropics/claude-for-legal9.6k2 repos~229Automated safety check: PassApache-2.0
Better Interfacetrycompai/crm11k—~2kAutomated safety check: PassMIT
Composition Patternssickn33/agentic-awesome-skills47k1 repos~742Automated safety check: PassMIT
Skill Gap Analysissickn33/agentic-awesome-skills47k1 repos~3.6kAutomated safety check: PassMIT

Similar skills

  • Diagnose

    github/awesome-copilot

    Official

    Perform a systematic diagnostic scan of an AI workflow across 5 quality dimensions — prompt quality, context efficiency, tool health, architecture fitness, and safety — producing a scored report…

    40k GitHub starsUsed in 1 repo~1k tokens
    Agent WorkflowsAuto-check passed
  • Gaps

    anthropics/claude-for-legal

    Official

    Open gaps tracker — what's flagged and not yet closed. An agent skill from anthropics/claude-for-legal.

    9.6k GitHub starsUsed in 2 repos~229 tokens
    Auto-check passed
  • Better Interface

    trycompai/crm

    User-invoked, cross-discipline interface review that coordinates better-accessibility, better-layout, better-writing, better-typography, better-colors, and better-ui.

    11k GitHub stars~2k tokensUpdated 26 days ago
    Frontend & DesignAuto-check passed
  • Composition Patterns

    sickn33/agentic-awesome-skills

    A skill your agent uses when working with composition-patterns tasks or workflows

    47k GitHub starsUsed in 1 repo~742 tokens
    DevelopmentAuto-check passed
  • Skill Gap Analysis

    sickn33/agentic-awesome-skills

    Skill gap register: employee, skill area, current against required level, gap severity and recommended training.

    47k GitHub starsUsed in 1 repo~3.6k tokens
    Auto-check passed
  • Diagnose Why Work Stopped

    paperclipai/paperclip

    Diagnose stalled, looping, or over-recovered Paperclip issue trees and propose a no-code product-rule plan.

    98k GitHub stars~2.8k tokensUpdated today
    Auto-check passed

More from manager-dot-dev/manager-skills

All 27 skills in this repo
  • Em Grid Scorer

    manager-dot-dev/manager-skills

    Score an Engineering Manager's coverage across all 12 cells of the EM Grid based on their calendar and Slack.

    114 GitHub stars~5.8k tokensUpdated 5 mo ago
    Auto-check passed
  • Em Context

    manager-dot-dev/manager-skills

    Foundation skill for engineering managers. An agent skill from manager-dot-dev/manager-skills.

    114 GitHub stars~1.6k tokensUpdated 5 mo ago
    Auto-check passed
  • 1on1s

    manager-dot-dev/manager-skills

    Prepares agendas, diagnoses struggling 1:1 relationships, and gives frameworks for running effective 1:1 meetings with direct reports.

    114 GitHub stars~3k tokensUpdated 5 mo ago
    Auto-check passed
  • Business Literacy

    manager-dot-dev/manager-skills

    Explains business financial terms and frameworks for engineering managers — produces term definitions (ARR, COGS, CAC, LTV, gross margin, burn rate, EBITDA, AARRR), translation formulas for making…

    114 GitHub stars~2.4k tokensUpdated 5 mo ago
    Auto-check passed
  • Career Development

    manager-dot-dev/manager-skills

    Helps engineering managers support direct report growth — produces a stage-by-stage model of engineering impact (Circles of Influence), a framework for non-linear career planning (Tarzan Method)…

    114 GitHub stars~2.3k tokensUpdated 5 mo ago
    Auto-check passed
  • Delegation

    manager-dot-dev/manager-skills

    Guides managers out of the bottleneck role — provides the Team Rep pattern, Epic Ownership model, Task-Relevant Maturity framework, kingdom ownership, and three-layer assignment strategy.

    114 GitHub stars~3.3k tokensUpdated 5 mo ago
    Auto-check passed

Questions about Team Composition

What does Team Composition do?

Helps engineering managers diagnose team skill gaps and make better hiring and assignment decisions — produces the Dungeon Party archetype model (Warrior, Tank, Healer, Wizard, Rogue), the Barrels…. Team Composition is an agent skill from manager-dot-dev/manager-skills. Helps engineering managers diagnose team skill gaps and make better hiring and assignment decisions — produces the Dungeon Party archetype model (Warrior, Tank, Healer, Wizard, Rogue), the Barrels and Ammunition framework for understanding throughput limits, the Commandos/Infantry/Police phase model, and a minimum team size guideline.

When should I use Team Composition?

Team Composition fits situations like: the user says team balance; what roles do I need; who should I hire next; team is missing something.

How do I install Team Composition in Claude Code?

Run `npx skills add manager-dot-dev/manager-skills --skill team-composition -a claude-code`. Or copy the skill folder (skills/team-composition in manager-dot-dev/manager-skills) into .claude/skills/team-composition in your project. Claude Code loads it when a task matches its description.

How do I install Team Composition in Codex?

Run `npx skills add manager-dot-dev/manager-skills --skill team-composition -a codex`. Or copy the skill folder (skills/team-composition in manager-dot-dev/manager-skills) into .agents/skills/team-composition in your project. Codex loads it when a task matches its description.

Can I use Team Composition in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add manager-dot-dev/manager-skills --skill team-composition -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/team-composition, .gemini/skills/team-composition, .github/skills/team-composition and .opencode/skills/team-composition in your project.

What does Team Composition need to run?

SKILL.md names no scripts, command-line tools or credentials: Team Composition is instructions for the agent only.

Does Team Composition access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Team Composition safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Team Composition use?

Team Composition is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Team Composition use?

About 2.5k tokens (SKILL.md is roughly 10k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 204 tokens, read only when the agent opens those files.

What are the alternatives to Team Composition?

Skills that share tags, products or a category with Team Composition: Diagnose (github/awesome-copilot, 40k stars), Gaps (anthropics/claude-for-legal, 9.6k stars), Better Interface (trycompai/crm, 11k stars) and Composition Patterns (sickn33/agentic-awesome-skills, 47k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Team Composition?

manager-dot-dev (a GitHub organization) maintains it in manager-dot-dev/manager-skills, which has 114 GitHub stars. The repository holds 27 skills in this directory. The repository was last updated on May 9, 2026.

Source: manager-dot-dev/manager-skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.