{ "markdown": "

\"logo\"

\n\n# claude-code-workflow-cheatsheet\n\n**Never forget a slash command again.** Auto-generated workflow cheatsheets that live inside Claude Code.\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)\n[![Claude Code](https://img.shields.io/badge/Claude_Code-Skill-blueviolet)](https://docs.anthropic.com/en/docs/claude-code)\n[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](CONTRIBUTING.md)\n\n---\n\n## The Problem\n\nYou installed [Superpowers](https://github.com/obra/superpowers), [Everything Claude Code](https://github.com/anthropics/everything-claude-code), [Oh My Claude Code](https://github.com/yeachan-heo/oh-my-claudecode), [Matt Pocock's skills](https://github.com/mattpocock/skills), or built your own commands — then forgot half of them exist. You end up using the same 3 commands while 40+ sit unused.\n\n## The Solution\n\nThis skill scans your installed commands and generates a **persistent cheatsheet** that Claude sees at the start of every session — via the `companyAnnouncements` field in `settings.json`.\n\n```\n┌─────────────────────────────────────────────────────────────────────┐\n│ What Claude sees every session: │\n│ │\n│ [Workflows] Dev: /orchestrate feature \"desc\" -> /e2e │\n│ [Workflows] Quality: /code-review -> /refactor-clean -> /verify │\n│ [Workflows] Multi: /multi-plan -> /multi-execute | /devfleet │\n│ [Workflows] Learn: /learn -> /learn-eval -> /skill-create │\n│ [OMC] Auto: autopilot: desc | Persist: ralph: desc │\n│ [Superpowers] Design: /brainstorming -> /writing-plans │\n│ ... │\n│ │\n│ Claude now proactively suggests the RIGHT command for your task. │\n└─────────────────────────────────────────────────────────────────────┘\n```\n\n**Before:** You type `/plan` for everything.\n**After:** Claude suggests `/orchestrate bugfix \"login crash\" -> /e2e` because it knows the full workflow.\n\n---\n\n## Quick Start\n\n### Option 1: Claude Code Skill (Recommended)\n\n```bash\n# Copy the skill into your Claude Code skills directory\ncp -r . ~/.claude/skills/company-announcements\n```\n\nThen in Claude Code:\n\n```\n/setup-announcements # Auto-detect your installed commands\n/setup-announcements --harness superpowers # Superpowers preset\n/setup-announcements --harness ecc # Everything Claude Code preset\n/setup-announcements --harness omcc # Oh My Claude Code preset\n/setup-announcements --harness mattpocock # Matt Pocock's engineering skills preset\n```\n\n### Option 2: Manual Install (30 seconds)\n\nPick a preset and paste it into `~/.claude/settings.json`:\n\n
\nSuperpowers (obra) - 5 workflow categories\n\n```json\n{\n \"companyAnnouncements\": [\n \"[Superpowers] Design: /brainstorming -> /writing-plans\\n[Superpowers] Execute: /using-git-worktrees -> (/executing-plans or /subagent-driven-development)\\n[Superpowers] Per Task: /test-driven-development -> /requesting-code-review -> /receiving-code-review\\n[Superpowers] Finish: /verification-before-completion -> /finishing-a-development-branch\\n[Superpowers] Tools: /systematic-debugging (bugs) | /dispatching-parallel-agents (parallel) | /writing-skills (meta)\"\n ]\n}\n```\n\n
\n\n
\nEverything Claude Code (ECC) - 7 workflow categories\n\n```json\n{\n \"companyAnnouncements\": [\n \"[Workflows] Dev: /orchestrate feature|bugfix \\\"desc\\\" -> /e2e | Manual: /plan -> /tdd -> /code-review -> /verify\\n[Workflows] Reproduce: /e2e (as-is) -> /orchestrate bugfix -> /e2e (to-be) | Build: /build-fix -> /verify\\n[Workflows] Quality: /code-review -> /refactor-clean -> /verify | /quality-gate | /eval define|check|report\\n[Workflows] Multi: /multi-plan -> /multi-execute | /devfleet \\\"task\\\" | /model-route \\\"task\\\" --budget low|med|high\\n[Workflows] Docs: /update-docs, /update-codemaps | Ref: /docs \\\"lib\\\" | Loop: /loop-start sequential|continuous-pr\\n[Workflows] Learn: /learn -> /learn-eval -> /skill-create | Instincts: /instinct-status -> /evolve -> /promote\\n[Workflows] Session: /save-session, /resume-session | Meta: /harness-audit, /skill-health, /context-budget\"\n ]\n}\n```\n\n
\n\n
\nOh My Claude Code (OMCC) - 6 workflow categories\n\n```json\n{\n \"companyAnnouncements\": [\n \"[OMC] Auto: autopilot: desc | Persist: ralph: desc | Parallel: ulw tasks\\n[OMC] Team: /oh-my-claudecode:team 3:executor \\\"task\\\" | Plan: ralplan | QA: /oh-my-claudecode:ultraqa\\n[OMC] Clarify: deep-interview \\\"idea\\\" -> omc-plan --consensus -> autopilot: desc\\n[OMC] Investigate: /oh-my-claudecode:trace | Tri-Model: ccg \\\"query\\\" | Visual: /oh-my-claudecode:visual-verdict\\n[OMC] Cleanup: deslop | Skills: /oh-my-claudecode:skill list | Learn: /oh-my-claudecode:learner\\n[OMC] Session: /oh-my-claudecode:psm | Release: /oh-my-claudecode:release | Diag: /oh-my-claudecode:omc-doctor\"\n ]\n}\n```\n\n
\n\n
\nVanilla Claude Code - Built-in commands only\n\n```json\n{\n \"companyAnnouncements\": [\n \"[Workflows] Dev: /plan -> /code-review | Bug: /build-fix\\n[Workflows] Docs: /docs \\\"lib\\\" | Session: /save-session, /resume-session\"\n ]\n}\n```\n\n
\n\n
\nMatt Pocock's skills - 6 disciplined workflows (plan, bug, feature, triage, explore)\n\n```json\n{\n \"companyAnnouncements\": [\n \"[Matt] Plan: /grill-with-docs -> /to-prd -> /to-issues (interview against CONTEXT.md, synthesize PRD, split into vertical-slice tracer-bullet issues)\\n[Matt] Bug: /diagnose (feedback loop -> reproduce -> 3-5 ranked hypotheses -> instrument one variable -> fix + regression test -> cleanup + post-mortem)\\n[Matt] Feature: /tdd (vertical tracer bullets one test -> one impl; NEVER write all tests then all code) | /prototype (throwaway: LOGIC terminal app or UI variants)\\n[Matt] Triage: /triage — sort a new/unsorted issue into needs-info (ask reporter) / ready-for-agent (AFK can grab it) / ready-for-human (judgment call) / wontfix. Use when issues pile up or before handing one to an agent.\\n[Matt] Explore: /zoom-out (unfamiliar code -> module + caller map) | /improve-codebase-architecture (shallow -> deep modules; HTML report in $TMPDIR)\"\n ]\n}\n```\n\nInstall Matt's skills first: `npx skills add https://github.com/mattpocock/skills`. Run `/setup-matt-pocock-skills` once per repo before first use — this configures the issue tracker (GitHub/GitLab/local), maps your 5 triage labels, and locates `CONTEXT.md`/`docs/adr/`.\n\nFor the mental model, skill-authoring rules, and a full quick-reference of Matt's canonical skills, see **[docs/mattpocock-cheatsheet.md](docs/mattpocock-cheatsheet.md)**.\n\n
\n\n---\n\n## How It Works\n\nClaude Code's `companyAnnouncements` field in `settings.json` is designed for team-wide messages. This skill repurposes it as a **persistent workflow memory** — a cheatsheet that Claude reads at the start of every session.\n\n> **Important:** `companyAnnouncements` must be a `string[]` (array). If the array has **multiple items**, Claude Code picks **one at random per session** (rotating banner). To show all lines every session, use a **single array item** with `\\n` for line breaks.\n\n```\nsettings.json Claude's System Prompt\n┌──────────────────────────┐ ┌──────────────────────────┐\n│ \"companyAnnouncements\": [ │ ────> │ Company announcements: │\n│ \"line1\\nline2\\nline3\" │ │ line1 │\n│ ] │ │ line2 │\n└──────────────────────────┘ │ line3 │\n └──────────────────────────┘\n │\n v\n Claude proactively suggests\n the right workflow chain\n```\n\n### Syntax Convention\n\n| Symbol | Meaning | Example |\n|--------|---------|---------|\n| `->` | Sequential steps | `/plan -> /tdd -> /verify` |\n| `\\|` | Alternative commands | `/verify \\| /quality-gate` |\n| `,` | Related commands | `/save-session, /resume-session` |\n\n---\n\n## Supported Harnesses\n\n| Harness | Commands | Style | Preset |\n|---------|----------|-------|--------|\n| [Superpowers](https://github.com/obra/superpowers) | 14 skills | `/brainstorming`, `/writing-plans`, `/executing-plans` | `superpowers` |\n| [Everything Claude Code](https://github.com/anthropics/everything-claude-code) | 40+ | `/plan`, `/tdd`, `/verify` | `ecc` |\n| [Oh My Claude Code](https://github.com/yeachan-heo/oh-my-claudecode) | 20+ | `/oh-my-claudecode:*` + magic keywords | `omcc` |\n| [Matt Pocock's skills](https://github.com/mattpocock/skills) | 6 workflows | `/diagnose`, `/tdd`, `/to-prd`, `/to-issues`, `/triage` | `mattpocock` |\n| Vanilla Claude Code | Built-in | `/plan`, `/code-review` | `minimal` |\n| Custom | Your own | Any combination | `custom` |\n\n---\n\n## Create Your Own Preset\n\n1. Create a JSON file in `templates/`:\n\n```json\n{\n \"_comment\": \"My team's workflow\",\n \"companyAnnouncements\": [\n \"[MyTeam] Deploy: /build -> /test -> /deploy-staging -> /deploy-prod\\n[MyTeam] Review: /lint -> /code-review -> /security-scan -> /approve\"\n ]\n}\n```\n\n2. Each workflow line should be under ~120 characters\n3. Use `[Prefix]` to group related workflows\n4. Submit a PR to share with the community!\n\n---\n\n## Writing Effective Announcements\n\n**Do:**\n- Group by workflow phase (Dev, Quality, Docs, etc.)\n- Show the complete chain with `->` arrows\n- Include alternatives with `|` pipes\n- Keep each line scannable at a glance\n\n**Don't:**\n- List every flag and option (keep it high-level)\n- Exceed 7-10 announcement lines (Claude's attention is finite)\n- Mix harness styles in the same config\n\n---\n\n## Examples\n\n### Mixed Harness Setup\n\nUsing both ECC and OMCC? Combine presets:\n\n```json\n{\n \"companyAnnouncements\": [\n \"[ECC] Dev: /orchestrate feature \\\"desc\\\" -> /e2e -> /verify\\n[ECC] Quality: /code-review -> /refactor-clean -> /quality-gate\\n[OMC] Auto: autopilot: desc | Persist: ralph: desc\\n[OMC] Team: /oh-my-claudecode:team 3:executor \\\"task\\\"\"\n ]\n}\n```\n\n### Project-Specific Workflows\n\nAdd project-level workflows alongside harness defaults:\n\n```json\n{\n \"companyAnnouncements\": [\n \"[Project] PR: /plan -> /tdd -> /code-review -> gh pr create\\n[Project] Hotfix: /orchestrate bugfix \\\"desc\\\" -> /e2e -> git push\\n[Workflows] Quality: /code-review -> /refactor-clean -> /verify\"\n ]\n}\n```\n\n---\n\n## FAQ\n\n**Q: Does this slow down Claude Code?**\nA: No. `companyAnnouncements` is injected into the system prompt — it costs a few hundred tokens at most and adds zero latency.\n\n**Q: Can I use this without any harness?**\nA: Yes! Use the `minimal` preset for vanilla Claude Code, or create your own custom workflows.\n\n**Q: What happens if I have too many announcements?**\nA: Keep it under 10 lines. More than that dilutes Claude's attention. Focus on workflows you actually use.\n\n**Q: Does this work with Claude Code in VS Code / JetBrains?**\nA: Yes. `settings.json` is shared across all Claude Code interfaces — CLI, desktop app, IDE extensions.\n\n---\n\n## Contributing\n\nContributions welcome! The most valuable additions are:\n\n1. **New presets** for popular harnesses or team workflows\n2. **Auto-detection improvements** for command discovery\n3. **Documentation** and workflow recipes\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n---\n\n## Related Projects\n\n- [Superpowers](https://github.com/obra/superpowers) — Agentic skills framework & dev methodology\n- [Everything Claude Code](https://github.com/anthropics/everything-claude-code) — Comprehensive skill/command harness\n- [Oh My Claude Code](https://github.com/yeachan-heo/oh-my-claudecode) — Autonomous agent workflows\n- [Matt Pocock's skills](https://github.com/mattpocock/skills) — Disciplined engineering workflows (plan / bug / feature / triage / explore). See the [field guide](docs/mattpocock-cheatsheet.md).\n- [Claude Code](https://docs.anthropic.com/en/docs/claude-code) — Official documentation\n\n---\n\n## License\n\n[MIT](LICENSE)\n", "bytes": 12380, "sha": "34b07937ea4d8a1c8c7b0b58095161a6a6095341db8f58fb0ce6d256c7190f1a", "repo_slug": "cskwork/claude-code-workflow-cheatsheet", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/plg_cskwork_claude_code_workflow_cheatsheet_7a730c32/readme" }