Skip to content

SwarmKit

A specialist agent swarm for Claude Code, OpenCode and Antigravity, built from one source.

How it works

A task comes in.

It is classified. The tier sets how much ceremony it gets.

Routed to the specialist that owns the work: backend-specialist.

Quality gates run in parallel, only where relevant.

Fix a typo, T1: Edit directly; No plan, no gate.

Add an API route, T2: backend-specialist, security-auditor, release-tester.

Refactor auth, T3: explore, backend-specialist, security-auditor, code-proofreader.

  • T1 Trivial edits
  • T2 Contained domain work
  • T3 Cross-cutting work

The roster

33 agents, one cell each. Hover or focus a cell.

  • Core 1
  • Context and fast work 3
  • Domain specialists 19
  • Quality gates 4
  • Showroom 6

Numbers come from runs, or they do not appear.

ponytail diff
net: pending
benchmarks
pending
harness
bench: Antigravity (agy) gemini-3.8-flash-high. site: Claude Code.
build log
30 human interventions, 4 agent dispatches

Install

./install.sh --all
  • Claude Code --claude
  • OpenCode --opencode
  • Antigravity --agy
  • 33 agents
  • 27 skills
  • 11 MCP servers
  • 5 slash commands
  • --n8n
  • --cloudflare
  • --all
  • --free
  • --uninstall
  • --help

Reference

The detail behind the scroll. You write every agent, skill and rule once in core/; a small compiler turns them into each CLI's native format and the installer links the results into place.

01 / Install

Install

Clone the repo and run one script; it links everything into place.

git clone https://github.com/ha-lun/swarmkit.git
cd swarmkit
./install.sh --all

Anything it replaces is backed up first. Running it again changes nothing. --all is also the default when no flag is given.

Platforms 3

  • Claude Code

    --claude

    Specialists install as subagents. Tool limits are enforced by tool lists and guard hooks.

  • OpenCode

    --opencode

    Specialists install as agents. Tool limits are enforced by permission blocks.

  • Antigravity

    --agy

    Specialists ship as skills in the swarmkit plugin. Antigravity has no custom subagents, so limits are stated in each skill.

Installer flags 9

FlagWhat it does
--opencodeInstall OpenCode config
--agyInstall Antigravity (agy) Swarm config
--claudeInstall Claude Code Swarm config
--n8nConfigure local self-hosted n8n credentials (optional add-on, not in --all)
--cloudflareInstall Cloudflare skills and configure auth (optional add-on, not in --all)
--allInstall all agent configs (opencode, agy, claude)
--freeEnable free mode for OpenCode (uses default models, no keys)
--uninstallUninstall all configurations
--helpShow this help message

02 / How it routes

How it routes

A request is classified into a tier, explored if needed, executed, then checked.

Routing tiers

TierKind of taskWhat happens
T0Questions and reviewsAnswered directly.
T1Trivial editsTypos, renames, version bumps. Done directly, with no plan and no gate.
T2Contained domain workPlan, your approval, then execution by the domain specialist.
T3Cross-cutting workSame as T2, in an isolated git worktree.

T0 to T3 are routing tiers (how much ceremony a task gets). fast, standard and deep are model tiers (which model class runs an agent).

Routing: a request is classified, explored, executed, then checked by quality gates in parallel where the CLI allowsSteps: route (classify T0 to T3), then explore (read-only context brief), then execute (specialist does the work). Then, in parallel where the CLI allows and when relevant: security-auditor when auth, secrets or input handling changed; code-proofreader when the diff is large; release-tester when tests were not run.routeclassify T0 to T3exploreread-only context briefexecutespecialist does the workin parallel where the CLI allowssecurity-auditorauth, secrets or input handling changedcode-proofreaderthe diff is largerelease-testertests were not run
Explore is for non-trivial work or uncertain context. It can be skipped when the task is self-contained.

Example routes

Illustrative: the actual route depends on what the change touches.

  • Fix a typo

    T1
    1. Edit directly
    2. No plan, no gate
  • Add an API route

    T2
    1. Plan and approval
    2. Implement backend-specialist
    3. If input handling changed security-auditor
    4. If tests were not run release-tester
  • Refactor auth

    T3
    1. Isolated worktree
    2. If the codebase is unfamiliar explore
    3. Plan and approval
    4. Implement backend-specialist
    5. If auth changed security-auditor
    6. If the diff is large code-proofreader

Quality gates 3

Run after the work, in parallel where the CLI allows, and only when relevant. Worktrees are never merged or removed without an explicit instruction.

AgentRuns when
security-auditorauth, secrets or input handling changed
code-proofreaderthe diff is large
release-testertests were not run

03 / Roster

Roster

33 agents, grouped by the job they do.

Core 1

Plans and dispatches. Does not touch files.

  • lead-devdeep

    Primary orchestrator

Context and fast work 3

Read-only context, git, mechanical edits.

  • explorefast

    Read-only context-gathering pre-flight for the lead-dev swarm

  • git-specialistfast

    Git workflow specialist — commit/branch review (default) AND...

  • junior-devfast

    Junior dev for light, mechanical code edits that don't need a domain...

Domain specialists 19

Where T2 and T3 work lands.

  • android-capacitor-specialiststandard

    Android specialist for Capacitor apps with React + Vite

  • animation-specialiststandard

    Animation, 2D, and 3D specialist for web — Motion, GSAP, Anime.js...

  • backend-specialiststandard

    Backend specialist focused on API design, service boundaries...

  • blender-specialiststandard

    3D modeling, mesh generation, asset staging, geometry nodes, spatial...

  • db-specialistdeep

    Database specialist for schema design, migrations, query...

  • devops-specialiststandard

    DevOps specialist for CI/CD pipelines, infrastructure as code...

  • docker-specialiststandard

    Docker specialist for containerization, Dockerfiles, Compose stacks...

  • electron-specialiststandard

    Electron specialist for wrapping existing React + Vite web apps as...

  • frontend-specialiststandard

    Frontend specialist focused on production-ready UI implementation...

  • ios-capacitor-specialiststandard

    iOS specialist for Capacitor apps with React + Vite

  • linkedin-specialiststandard

    LinkedIn content specialist

  • lovable-specialiststandard

    Frontend specialist for Lovable-made projects

  • monitoring-specialiststandard

    Monitoring and observability specialist for Prometheus, Grafana...

  • n8n-debuggerstandard

    Systematic debugging and diagnosis of broken n8n workflows

  • n8n-workflow-builderstandard

    Build and design n8n workflows from requirements

  • seo-specialiststandard

    SEO specialist — makes sure websites actually get seen by Google and...

  • seo-workerstandard

    Post-build SEO & sharing pass worker

  • server-specialiststandard

    Ubuntu server administration expert for system configuration...

  • swarm-architectdeep

    Swarm architect specialist for designing, extending, and refactoring...

Quality gates 4

Review, verify and test.

  • code-proofreaderstandard

    Code proofreader that finds dead code, redundant logic, unused...

  • release-testerfast

    Final quality gate that runs tests, linters, type checkers, and...

  • security-auditordeep

    Security reviewer that scans code for secrets leakage, hardcoded API...

  • test-writerstandard

    Writes unit and integration tests for new code

Showroom 6

A coordinator and its workers for scroll-driven product pages.

  • showroom-art-directorstandard

    Showroom Art Director

  • showroom-asset-processorfast

    Showroom Asset Processor Worker

  • showroom-frontend-builderstandard

    Showroom Frontend Builder Worker

  • showroom-intakestandard

    Showroom Intake Worker

  • showroom-motion-engineerstandard

    Showroom Motion Engineer Worker

  • showroomstandard

    Showroom Coordinator

04 / Extend

Extend

MCP servers, skills and slash commands that ship with the kit.

MCP servers 11

Registered for Claude Code and Antigravity from core/mcp.json.

  • blender
  • cloudflare
  • cloudflare-bindings
  • cloudflare-builds
  • cloudflare-docs
  • cloudflare-observability
  • firecrawl
  • gemini-mcp-tool
  • google-search-console
  • google-trends
  • playwright

Skills 27

Shared by all three CLIs, from core/skills/.

  • backend-quality
  • capacitor-mobile-quality
  • caveman
  • caveman-commit
  • caveman-compress
  • caveman-help
  • caveman-review
  • curated-resources
  • frontend-quality
  • git-workflow
  • img2threejs
  • n8n-api
  • n8n-debugging
  • ponytail
  • ponytail-audit
  • ponytail-debt
  • ponytail-help
  • ponytail-review
  • premium-frontend-system
  • release-testing
  • scroll-craft
  • security-review
  • seo-engineering
  • seo-sharing-pass
  • showroom
  • swarm-handoff
  • web-design-guidelines

Slash commands 5

OpenCode.

  • /ponytail-auditAudit the whole repo for over-engineering, what can be deleted
  • /ponytail-debtHarvest ponytail: comments into a tracked debt ledger
  • /ponytail-helpQuick reference for ponytail levels, skills, and commands
  • /ponytail-reviewReview changes for over-engineering, what can be deleted
  • /ponytailSwitch ponytail intensity level (lite/full/ultra/off)

05 / Proof

Proof

What is measured, what is not yet, and how this site was made.

Ponytail diff

net: pending

Ponytail is the always-on discipline against over-engineering. No before-and-after diff has been recorded yet, so no figure is shown.

Benchmarks

benchmarks pending

No benchmark data has been recorded yet. When it exists it will show medians over at least three runs per fixture per agent type, with the raw file linked. Losses will be shown too.

Harness disclosure

Benchmark runs: Antigravity (agy) with gemini-3.8-flash-high. This site: built with Claude Code, using the agents described above. The scene is procedural: no external models, textures or volume data, and no AI-generated media.

06 / Build log

Build log

Who did what, read from the repo at build time. It separates what the human did from what agents did.

Human interventions
30
4 gate approvals, 18 decisions and answers, 8 art-direction rounds with a human verdict. Measurements and agent-applied changes are not counted.
Agent dispatches
4
animation-specialist 2, frontend-specialist 1, showroom 1
Build timeline 44 events
  1. Human decision

    Phase 0 benchmark deferred: user will supply benchmark file later

  2. Human decision

    Dev server binds to 0.0.0.0 (global rule overrides PLAN.md 127.0.0.1)

  3. Agent dispatch

    showroom finished a task

  4. Human gate

    BRIEF.md approved; five-chapter structure per PLAN section 6 confirmed

  5. Human decision

    Phase 2 plan approved: worktree site/previs, animation-specialist builds grey-box previs; go/no-go after

  6. Agent dispatch

    animation-specialist finished a task

  7. Agent art direction

    Round 1: 1 change applied by the agent

  8. Human measurement

    Phase 2 frame rate reported by the human

  9. Human gate

    grey-box previs approved; fps within budget on laptop and phone

  10. Human decision

    Phase 3 plan approved: ink cool blue-black; Space Grotesk + JetBrains Mono; accent = 3 candidates in round 1

  11. Agent dispatch

    animation-specialist finished a task

  12. Human art direction

    Round 2: 1 verdict from the human

  13. Agent art direction

    Round 2: 6 changes applied by the agent

  14. Human decision

    Human: focus on desktop, skip phone fps checks for now. Phone budget and medium tier are deferred, not dropped from PLAN.md

  15. Human gate

    human: it all looks good; all elements and tokens locked (violet accent, cool ink, Space Grotesk + JetBrains Mono). Desktop fps sweep not reported; phone/medium tier deferred

  16. Human art direction

    Round 3: 6 verdicts from the human

  17. Human decision

    Phase 4 plan approved: ponytail diff + benchmarks pending; repo github.com/ha-lun/swarmkit; horology placeholder; responsive at 1440/1024/768/390 layout only

  18. Agent dispatch

    frontend-specialist finished a task

  19. Human gate

    human: the page looks good; continue. Review items 1-5 (interventions count definition, parallel-gates qualifier, terse roster blurbs, tier legend, benchmark schema) not yet answered

  20. Human decision

    Phase 5 plan approved; interventions counter counts only the human own actions

  21. Human decision

    Human feedback after Phase 5: animation should be the centre, small captions only. Plan 5b approved: detail text moves into the world (cell label cards) plus a Reference section; pinned scenes; slim Proof HUD; finale is the climax

  22. Human decision

    Human: animation-first build is better; wants the honeycomb as a sphere/globe. Plan 5c approved: Goldberg sphere, scroll orbits camera, satellites as orbiting moon cluster; new art-direction round 4 to follow

  23. Human decision

    Human: yes to round-4 tuning (finer globe 252 cells, lower columns, softer rim, pulled-back fly-over). Material = pearlescent ceramic; surroundings = atmosphere halo + sparse dust + moving studio light; text = captions 16px, large wordmark/tagline at start. Look values unlocked for round 4 only

  24. Human art direction

    Round 4: 1 verdict from the human

  25. Agent art direction

    Round 4: 1 change applied by the agent

  26. Human decision

    Human round-4 feedback: dislikes the glow (all of it removed), comet animation not up to par, transitions not smooth, globe has no texture and panels misaligned. Chose speckled stone surface and remove all glow. Round 5 plan approved

  27. Human decision

    Human after round 5: still some work to do but very good. Asked to commit everything and resume in a new session. Round-5 look NOT locked; site/integration NOT merged

  28. Human art direction

    Round 5: 2 verdicts from the human

  29. Human decision

    Round 6 plan approved: elevation terrain + rock shader, polished cut facets replace rings, seamless cells-proof dissolve (live-pose outgoing frame, post stays on, one ease), scene text fixed and fade-only. Look values remain unlocked

  30. Human art direction

    Round 6: 1 verdict from the human

  31. Agent art direction

    Round 6: 1 change applied by the agent

  32. Human decision

    Round 7 plan approved: basalt columns (quantised stepped heights, flat smooth tops, dark sides), palette tokens UNLOCKED for this round (graphite, iron-blue, bone candidates via ?palette=), accent stays violet. Human asked whether an AI-generated image would improve the texture, then said go ahead with the plan as written: no AI media used

  33. Human art direction

    Round 7: 1 verdict from the human

  34. Agent art direction

    Round 7: 1 change applied by the agent

  35. Human decision

    Palette LOCKED to iron-blue (ink #080b10, ink-2 #121824, wax #9aa6b4, wax-dim #46505e, text #d9dee5; accent stays violet). Round 8 plan approved: piston columns, slow spin parked during hive and fly-over, agents steady

  36. Human art direction

    Round 8: 1 verdict from the human

  37. Agent art direction

    Round 8: 1 change applied by the agent

  38. Human decision

    Round 9 plan approved: agents as fixed towers above the moving rods with polished caps; richer top and wall shading and lighting; key-light cast shadows on high tier only (medium gets AO stand-in). Palette stays locked (iron-blue)

  39. Human art direction

    Round 9: 1 verdict from the human

  40. Agent art direction

    Round 9: 1 change applied by the agent

  41. Human decision

    Round 10 plan approved: moon spin, drift and light piston pulse parked home for hive and fly-over; Reference redesigned as index rail plus six focused sections (Install, How it routes, Roster, Extend, Proof, Build log), honest pending states kept

  42. Agent art direction

    Round 10: 1 change applied by the agent

  43. Human decision

    Human ended the session and will continue tomorrow. Round 10 (moon motion, Reference rebuild) built, no verdict yet; look values still unlocked; site/integration NOT merged; HANDOFF.md updated

  44. Agent art direction

    Round 11: 1 change applied by the agent

Horology demo: link coming.Source: github.com/ha-lun/swarmkit