~ / rigs / eduardopin / agenfk-squad

eduardopin/agenfk-squad

A squad of Claude Code agents and a watcher hook that keeps parallel AI sessions honest: context budgets, dispatch limits, independent review and a merge queue. Built on AgenFK.

↗ GitHub ★ 0 NOASSERTION updated 11d ago project Claude CodeCodex
share on X
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
CONTEXT TAX · EVERY TURN
~6.0k tokens
Moderate · median rig: 2.3k · breakdown
GUARDRAILS
0/5
No committed guardrails · details
Quality 3/100 Guardrails v2

Copy this rig

# review before running: this installs third-party code
$ npx degit eduardopin/agenfk-squad/agents ./rig-agenfk-squad/agents
$ npx degit eduardopin/agenfk-squad/skills ./rig-agenfk-squad/skills
$ npx degit eduardopin/agenfk-squad/hooks ./rig-agenfk-squad/hooks

MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.

$ npx degit eduardopin/agenfk-squad/skills/agent-orchestration .claude/skills/agent-orchestration
$ npx degit eduardopin/agenfk-squad/skills/architect .claude/skills/architect
$ npx degit eduardopin/agenfk-squad/skills/board-heartbeat .claude/skills/board-heartbeat
$ npx degit eduardopin/agenfk-squad/skills/ci-cost .claude/skills/ci-cost
$ npx degit eduardopin/agenfk-squad/skills/ci-plan-evidence .claude/skills/ci-plan-evidence
$ npx degit eduardopin/agenfk-squad/skills/clear-pickup .claude/skills/clear-pickup
$ npx degit eduardopin/agenfk-squad/skills/cloud-diagram .claude/skills/cloud-diagram
$ npx degit eduardopin/agenfk-squad/skills/cloud-resource .claude/skills/cloud-resource
$ npx degit eduardopin/agenfk-squad/skills/design-bridge .claude/skills/design-bridge
$ npx degit eduardopin/agenfk-squad/skills/design-reference .claude/skills/design-reference
$ npx degit eduardopin/agenfk-squad/skills/engineering-evidence .claude/skills/engineering-evidence
$ npx degit eduardopin/agenfk-squad/skills/graph-context .claude/skills/graph-context
$ npx degit eduardopin/agenfk-squad/skills/harness-port .claude/skills/harness-port
$ npx degit eduardopin/agenfk-squad/skills/journey-qa .claude/skills/journey-qa
$ npx degit eduardopin/agenfk-squad/skills/parallel-git .claude/skills/parallel-git
$ npx degit eduardopin/agenfk-squad/skills/parallel-plan .claude/skills/parallel-plan
$ npx degit eduardopin/agenfk-squad/skills/pr-evidence .claude/skills/pr-evidence
$ npx degit eduardopin/agenfk-squad/skills/presentation .claude/skills/presentation
$ npx degit eduardopin/agenfk-squad/skills/project-rails .claude/skills/project-rails
$ npx degit eduardopin/agenfk-squad/skills/pull-request .claude/skills/pull-request
$ npx degit eduardopin/agenfk-squad/skills/session-context .claude/skills/session-context
$ npx degit eduardopin/agenfk-squad/skills/skill-trial .claude/skills/skill-trial
$ npx degit eduardopin/agenfk-squad/skills/writing .claude/skills/writing
$ curl -fsSL --create-dirs -o .claude/agents/context.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/context.md
$ curl -fsSL --create-dirs -o .claude/agents/design-bridge.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/design-bridge.md
$ curl -fsSL --create-dirs -o .claude/agents/developer.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/developer.md
$ curl -fsSL --create-dirs -o .claude/agents/finance.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/finance.md
$ curl -fsSL --create-dirs -o .claude/agents/finops.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/finops.md
$ curl -fsSL --create-dirs -o .claude/agents/frontend.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/frontend.md
$ curl -fsSL --create-dirs -o .claude/agents/github.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/github.md
$ curl -fsSL --create-dirs -o .claude/agents/gtm.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/gtm.md
$ curl -fsSL --create-dirs -o .claude/agents/infra.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/infra.md
$ curl -fsSL --create-dirs -o .claude/agents/product.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/product.md
$ curl -fsSL --create-dirs -o .claude/agents/tester.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/tester.md
$ curl -fsSL --create-dirs -o .claude/agents/verifier.md https://raw.githubusercontent.com/eduardopin/agenfk-squad/beta/agents/verifier.md

This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:

{
  "permissions": {
    "deny": [
      "Read(./.env)",
      "Read(./.env.*)",
      "Read(**/.env)",
      "Read(~/.ssh/**)",
      "Bash(git push --force:*)",
      "Read(**/*.pem)",
      "Read(.env)",
      "Read(~/.aws/**)",
      "Read(**/*.key)",
      "Read(**/.env.*)",
      "Bash(git push --force*)",
      "Bash(rm -rf:*)",
      "Bash(rm -rf *)",
      "Bash(sudo *)",
      "Bash(sudo:*)",
      "Bash(rm -rf /)",
      "Bash(git push -f:*)",
      "Read(.env.*)",
      "Bash(git reset --hard*)",
      "Bash(rm -rf /*)"
    ],
    "ask": [
      "Bash(git push:*)",
      "Bash(git push *)",
      "Bash(gh pr merge *)",
      "Bash(git commit:*)",
      "Bash(rm *)",
      "Bash(gh pr merge:*)",
      "Bash(gh pr create *)",
      "Bash(git reset --hard:*)",
      "Bash(npm publish:*)",
      "Bash(git rebase *)"
    ]
  }
}

Skills (23)

Subagents (12)

context
model: sonnet
Context manager for AI sessions (Claude Code, Codex, pi). Measures context, weighted burn, reread, waste, loops and handoff age for every live session, decides between continuing, compacting, handing
design-bridge
model: inherit
The squad's specialist for carrying design between code, Figma, the Claude Design System, GPT (image and video on the owner's ChatGPT plan) and local ComfyUI. Mirrors motion from code into Figma with
developer
model: inherit
Implements exactly ONE AgenFK card, on its own branch and in its own worktree, with the project's suite as the gate. Use proactively whenever a card with clear scope and acceptance criteria needs to b
finance
model: inherit
Product finance, read-only. Unit economics, pricing, margin, break-even and runway for a project or for the portfolio, always from sourced data. Use when the request involves pricing, a paid plan, mar
finops
model: inherit
Engineering cost for any project, read-only. Quotes the monthly cloud cost of a design (line by line, with source, date and formula, at the measured volume and at 10×) and measures token/LLM spend per
frontend
model: inherit
The squad's design director and design engineer. Creates, rebuilds and audits interfaces that become a reference (web, mobile, desktop, dashboard, landing page, app shell, component, design system, to
github
model: inherit
The squad's controller of branches, worktrees, PRs and parallelism in any project. Maps who is in which worktree and branch (Claude Code, Codex and pi sessions, AgenFK cards, open PRs), plans parallel
gtm
model: inherit
Sales and go-to-market for any project, drafts only. Positioning, ideal customer, competitor analysis, value proposition, landing page and pricing page copy, outreach and demo scripts, channel list. U
infra
model: inherit
Infrastructure, CI/CD and operations for any project (AWS, OCI, Alibaba, Cloudflare, GitHub, docker). Writes and validates Terraform and pipelines and STOPS AT PLAN; never apply, destroy or deploy. Al
product
model: inherit
Product and management for any project, read-only on the code. Says where the project REALLY is (repo and board, not the plan on paper), measures the distance to a real user, cuts the goal down to the
tester
model: sonnet
Writes and runs the tests for ONE AgenFK card (happy path, edge cases and failures) and records the result on the card. Use proactively after every implementation, when a defect found by running the s
verifier
model: sonnet
INDEPENDENT reviewer and verifier, read-only, in a clean context. Judges the branch of an AgenFK card against the acceptance criteria, runs the suite on its own, checks each acceptance criterion again

Similar rigs

copied ✓