~ / rigs / heymegabyte / agent-skills

heymegabyte/agent-skills

Agent Skills — agent-neutral autonomous product-building OS for 32+ AI coding tools. 23 skill categories · 28 agents · 149 reference docs · 37 platform variants. One-line prompts → deployed products on Cloudflare Workers.

↗ GitHub ★ 22 other updated 4d ago project Claude CodeCodexCursorGemini CLICopilot Claude plugin
share on X
ARCHETYPE
YOLO Cowboy
Permissions? Never heard of 'em. Ships at the speed of `--dangerously-skip-permissions`.
CONTEXT TAX · EVERY TURN
~8.8k tokens
Heavy · median rig: 2.2k · breakdown
GUARDRAILS
3/5
Blocks destructive commands · Pre-tool screening hook · Sandbox or ask-first rules · details

Copy this rig

# review before running: this installs third-party code
$ claude mcp add --transport http cloudflare-api https://mcp.cloudflare.com/mcp
$ npx degit heymegabyte/agent-skills/agents ./rig-agent-skills/agents
$ npx degit heymegabyte/agent-skills/commands ./rig-agent-skills/commands

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.

$ claude mcp add --transport http cloudflare-api https://mcp.cloudflare.com/mcp
$ npx degit heymegabyte/agent-skills/.devin/skills/emdash .claude/skills/emdash
$ curl -fsSL --create-dirs -o .claude/agents/accessibility-auditor.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/accessibility-auditor.md
$ curl -fsSL --create-dirs -o .claude/agents/architect.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/architect.md
$ curl -fsSL --create-dirs -o .claude/agents/browser-operator.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/browser-operator.md
$ curl -fsSL --create-dirs -o .claude/agents/changelog-drafter.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/changelog-drafter.md
$ curl -fsSL --create-dirs -o .claude/agents/changelog-generator.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/changelog-generator.md
$ curl -fsSL --create-dirs -o .claude/agents/code-simplifier.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/code-simplifier.md
$ curl -fsSL --create-dirs -o .claude/agents/completeness-checker.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/completeness-checker.md
$ curl -fsSL --create-dirs -o .claude/agents/computer-use-operator.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/computer-use-operator.md
$ curl -fsSL --create-dirs -o .claude/agents/content-writer.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/content-writer.md
$ curl -fsSL --create-dirs -o .claude/agents/cost-estimator.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/cost-estimator.md
$ curl -fsSL --create-dirs -o .claude/agents/dead-code-remover.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/dead-code-remover.md
$ curl -fsSL --create-dirs -o .claude/agents/dependency-auditor.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/dependency-auditor.md
$ curl -fsSL --create-dirs -o .claude/agents/deploy-verifier.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/deploy-verifier.md
$ curl -fsSL --create-dirs -o .claude/agents/formatter.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/formatter.md
$ curl -fsSL --create-dirs -o .claude/agents/incident-responder.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/incident-responder.md
$ curl -fsSL --create-dirs -o .claude/agents/media-orchestrator.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/media-orchestrator.md
$ curl -fsSL --create-dirs -o .claude/agents/meta-orchestrator.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/meta-orchestrator.md
$ curl -fsSL --create-dirs -o .claude/agents/migration-agent.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/migration-agent.md
$ curl -fsSL --create-dirs -o .claude/agents/model-router.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/model-router.md
$ curl -fsSL --create-dirs -o .claude/agents/motion-choreographer.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/motion-choreographer.md
$ curl -fsSL --create-dirs -o .claude/agents/performance-profiler.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/performance-profiler.md
$ curl -fsSL --create-dirs -o .claude/agents/renamer.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/renamer.md
$ curl -fsSL --create-dirs -o .claude/agents/resource-broker.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/resource-broker.md
$ curl -fsSL --create-dirs -o .claude/agents/security-reviewer.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/security-reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/seo-auditor.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/seo-auditor.md
$ curl -fsSL --create-dirs -o .claude/agents/test-writer.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/test-writer.md
$ curl -fsSL --create-dirs -o .claude/agents/transcriber.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/transcriber.md
$ curl -fsSL --create-dirs -o .claude/agents/visual-qa.md https://raw.githubusercontent.com/heymegabyte/agent-skills/master/agents/visual-qa.md

Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.

{
  "permissions": {
    "deny": [
      "Bash(rm -rf /)",
      "Bash(rm -rf /*)",
      "Bash(rm -rf ~)",
      "Bash(rm -rf $HOME)",
      "Bash(> /dev/sda*)",
      "Bash(mkfs*)",
      "Bash(dd if=*of=/dev/*)",
      "Bash(chmod -R 777 /)"
    ]
  },
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Bash",
        "hooks": [
          {
            "type": "command",
            "command": "BLOCKED=$(echo \"$CLAUDE_TOOL_INPUT\" | grep -iE 'rm -rf[[:space:]]*/|rm -rf[[:space:]]*~/|rm -rf[[:space:]]*\\$HOME|git push --force.*main|git push -f.*main|git reset --hard|DROP TABLE|DROP DATABASE|DROP SCHEMA|TRUNCATE TABLE' 2>/dev/null); i"
          },
          {
            "type": "prompt",
            "command": "Is this bash command destructive to filesystem, database, or git history in a way that cannot be easily reversed? Only answer 'BLOCK' if genuinely dangerous. Command: $ARGUMENTS"
          }
        ]
      }
    ]
  },
  "sandbox": {
    "enabled": true
  }
}

MCP servers (1)

serversourceest. tokens
Cloudflare · "cloudflare-api" remote · remote 4.0k

Skills (1)

Subagents (28)

accessibility-auditor
model: sonnet
Dedicated axe-core + Playwright accessibility agent. Navigates pages, runs WCAG 2.2 AA audits at 6 breakpoints, reports violations with fix suggestions, verifies remediation.
architect
model: opus
Pre-implementation architecture agent. Analyzes project structure, generates repo-map, designs task graph, identifies architectural seams and parallel workstreams.
browser-operator
model: sonnet
Drives a REAL browser to verify deployments + operate the product like a user. Use after any deploy, for golden-path/visual/console/network verification, /admin inspection, and read-only account-resou
changelog-drafter
model: haiku
Reads git log since last tag, drafts CHANGELOG entry. Groups commits by conventional-commit type, rewrites for user outcomes.
changelog-generator
model: haiku
Auto-generates changelogs from conventional commits. Parses git log since last tag, groups by type, writes user-outcome-focused CHANGELOG.md entries.
code-simplifier
model: haiku
Simplifies code for clarity, consistency, and maintainability. Reduces complexity, flattens nesting, removes dead code, consolidates duplicates. Focuses on recently modified files.
completeness-checker
model: opus
Post-implementation verification agent. Runs Feature Completeness Engine, Zero Recommendations Gate, and visual verification. Only declares DONE when genuinely complete.
computer-use-operator
model: opus
Desktop automation specialist. Controls native macOS apps via Computer Use MCP. Handles Finder, System Settings, Preview, Notes, and cross-app workflows.
content-writer
model: sonnet
Marketing copy, blog posts, and SEO content writer using Emdash brand voice. Flesch>=60, active voice, sharp and punchy — zero AI slop.
cost-estimator
model: haiku
Estimates Cloudflare Workers costs before deploying. Reads wrangler.toml, counts D1 tables/rows, estimates KV/R2 usage, calculates monthly cost, warns on free tier limits.
dead-code-remover
model: haiku
Finds unreachable exports, unused imports, orphaned functions, and dead variables. Greps project to confirm no references, then removes safely.
dependency-auditor
model: haiku
Scans packages for outdated deps, security advisories, license violations, and unused imports. Proposes version bumps, runs tests after updates, generates prioritized upgrade report.
deploy-verifier
model: sonnet
Post-deploy smoke test agent. Verifies live URL, checks console errors, screenshots 6 breakpoints, runs axe-core, validates SEO tags, reports pass/fail with evidence.
formatter
model: haiku
Runs prettier + oxlint --fix on modified files. Ensures consistent formatting and lint compliance with zero human intervention.
incident-responder
model: opus
Sentry-triggered incident response agent. Reads error events, traces to source code, proposes fixes, creates branches, and opens PRs with the fix.
media-orchestrator
model: sonnet
Creates, optimizes, and orchestrates media assets — images (ResGrid, AVIF/WebP), video (clips, GIFs, Veo), audio (TTS, podcast), and 3D. Coordinates Cloudflare Images, R2, and AI pipeline.
meta-orchestrator
model: opus
Master agent coordinating all tools, MCPs, agents, and skills. Plans multi-system workflows by chaining MCPs, spawning specialized agents, and using Computer Use for native apps.
migration-agent
model: sonnet
Drizzle schema migration agent. Generates migrations from schema diffs, validates against D1/SQLite, tests rollbacks, ensures zero-downtime compatibility.
model-router
model: haiku
Use when dispatching work and unsure which model tier to use — returns a recommended model and effort level based on task characteristics
motion-choreographer
model: sonnet
CSS animation and motion designer. Implements View Transitions API, scroll-driven animations, CSS keyframes, FLIP, staggered reveals, and reduced-motion respected animations.
performance-profiler
model: opus
Runs Lighthouse audits, analyzes Core Web Vitals, suggests specific fixes with file:line references. Targets the house cinematic bar (LCP<=2.0s, CLS<=0.05, INP<=100ms per _kernel/standards.md#cwv); cl
renamer
model: haiku
Semantic rename across codebase. Greps all references, updates imports and usages, renames file or symbol. Safe rename with verification.
resource-broker
model: sonnet
Maintains a normalized registry of legitimately-owned account resources (credits, quotas, free tiers, promos, expiring entitlements) with secret-REFERENCES only, and routes work cost-aware. Use during
security-reviewer
model: opus
OWASP Top 10 security auditor. Reviews for injection flaws, secrets exposure, auth bypasses, CSP issues, and vulnerable dependencies. Read-only — never modifies code.
seo-auditor
model: sonnet
Audits pages for SEO compliance — title, meta, H1, JSON-LD, OG tags, internal links, sitemap, robots.txt. Uses Playwright to check live pages.
test-writer
model: sonnet
TDD-first test engineer. Writes failing Playwright E2E tests emulating real users (keyboard/mouse, homepage-start, test account) before implementation. Vitest for units.
transcriber
model: haiku
Transcribes a plan step into code, one file at a time. Reads the spec, generates the implementation, writes clean production code with no oversight.
visual-qa
model: opus
Visual quality assurance agent. Screenshots pages at all breakpoints, uses AI vision to detect layout breaks, misalignment, text overflow, broken images, and design inconsistencies.

Hooks (4)

eventmatcherruns
PreToolUseBashBLOCKED=$(echo "$CLAUDE_TOOL_INPUT" | grep -iE 'rm -rf[[:space:]]*/|rm -rf[[:space:]]*~/|rm -rf[[:space:]]*\$HOME|git push --force.*main|git push -f.*main|git reset --hard|DROP TABLE|DROP DATABASE|DROP SCHEMA|TRUNCATE TABLE' 2>/dev/null); i
PreToolUseBashIs this bash command destructive to filesystem, database, or git history in a way that cannot be easily reversed? Only answer 'BLOCK' if genuinely dangerous. Command: $ARGUMENTS
PreCompact*python3 ~/.claude/hooks/precompact-snapshot.py
Setup*mkdir -p .claude/plans .claude/rules && echo '{"setup":true}' > /dev/null

Slash commands (36)

/agent-diversity-review/bootstrap-status/claude-bootstrap/cloudflare-bootstrap/create-specialist-agent/customer-changelog-check/dashboard-cockpit/deploy/drift-check/fast-fix/final-review/improve-lint/install-lint-stack/list-arcs/multimedia-enrich/plan-execute-verify-repair/post-arc-retrospective/prepare-multi-file-brief/prepare-skeleton-brief/process/retro/review-global-prompts/rollback/run-evals/run-the-loop/saas/security-supply-chain/self-improve/session-recap/ship/skill-health/skip-questions/sync-claude/update-agentskills/vendor-rotation-calendar/verify-production

Plugins (8)

claude-md-management@claude-plugins-officialcode-review@claude-plugins-officialcommit-commands@claude-plugins-officialcode-simplifier@claude-plugins-officialfrontend-design@claude-plugins-officialhookify@claude-plugins-officialfeature-dev@claude-plugins-officialcloudflare@cloudflare

Cursor rules (1)

emdash-skills · always

Permissions

deny (8)
Bash(rm -rf /)
Bash(rm -rf /*)
Bash(rm -rf ~)
Bash(rm -rf $HOME)
Bash(> /dev/sda*)
Bash(mkfs*)
Bash(dd if=*of=/dev/*)
Bash(chmod -R 777 /)
ask (0)
—
allow (14)
Read
Write
Edit
Glob
Grep
Bash
WebFetch
WebSearch
Skill(*)
Agent(*)
mcp__*
NotebookEdit
EnterWorktree
ExitWorktree

Similar rigs

copied ✓