⚡ Skills Catalog
All 50 reusable commands across HIVE. Invoke any skill with /skill-name inside Claude Code.
50 total skills
·
10 squads
·
7 multi-model (OpenRouter)
5
Global
7
OpenRouter
4
Orchestrator
4
Commercial
8
Dev
11
Marketing
Global Skills
Available in every squad context — from root or any squad folder
| Skill | Command | What it does |
|---|---|---|
| HIVE Setup | /hive-setup | Guided onboarding (~10 min) — fills all squad contexts with your company name, industry, working style, and preferences |
| Open Squad | /open-squad <name> | git pull + load CLAUDE.md + STATE.md; aggregates L1 if squad has children |
| Close Squad | /close-squad <name> | Update STATE.md + decisions log + propagate L1 to parent orchestrator |
| Company Status | /status | Aggregate L1 from all active squads in one view — what's happening now across the company |
| Log Incident | /log-incident | Create structured incident record in incidents/ hub with root cause, fix, and prevention notes |
OpenRouter Skills
Multi-model AI queries — Kimi, GPT-5, Grok, Gemini, DeepSeek and more
Setup required: Set
OPENROUTER_API_KEY env var, then run bash _core/mcp/install.sh. See MCP setup guide for details. Model responses are always delivered raw — no editing, no paraphrasing.
| Skill | Command | What it does | Est. cost |
|---|---|---|---|
| Second Opinion | /second-opinion | Single model query — raw response, zero editing. Asks for model if not specified. | $0.001–$0.05 |
| Consensus | /consensus | Same prompt to N models in parallel. Shows raw responses + overlap/divergence synthesis. Optional --deep mode for cross-critique. |
$0.01–$0.20 |
| Anti-Bubble | /anti-bubble | Neutral model reads copy/pitch with NO company context — pure outsider reaction. Finds jargon and assumptions insiders can't see. | $0.005–$0.05 |
| ICP Check | /icp-check | 2–3 models simulate your ICP reading a piece of copy, email, or landing page. Cross-cultural triangulation. Reads foundation/icp-audience.md or inline description. |
$0.01–$0.15 |
| DoD Check | /dod-check | Cheap model validates PR diff vs Linear acceptance criteria. Extracts issue from branch name. Returns READY / ADJUSTMENTS / REDO verdict. | $0.001–$0.01 |
| Cost Compare | /cost-compare | Same task across 3–6 models from cheapest to top tier. Real token cost matrix. Helps find the "cheap enough" model for recurring workers. | $0.01–$0.50 |
| Code Review (model) | /code-review-model | Code review from any OpenRouter model. Same P1/P2/P3 classification as /codex-review. Use as second opinion or security scan. |
$0.01–$0.10 |
Model quick reference
| Alias | Model | Best for | Approx price |
|---|---|---|---|
| kimi | Kimi K2.6 | Coding, top open-weight | $0.68/$3.42 per M |
| gpt5.5 | GPT-5.5 | Reasoning, flagship | premium |
| grok | Grok 4.20 | 2M context, low cost | $0.30/$0.50 per M |
| gemini-pro | Gemini (latest) | Multimodal, long context | $0.25/$0.75 per M |
| dsflash | DeepSeek Flash | Near-free, 1M ctx | $0.10/$0.20 per M |
| qwenflash | Qwen Flash | Fast, cheap | $0.19/$1.13 per M |
| granite | IBM Granite | Cheapest tier | $0.05/$0.10 per M |
Full catalog (30+ models) with benchmarks and context windows: _core/MODELS-MAP.md
Orchestrator Skills
Stamper — Chief of Staff · Issue and project lifecycle management
/create-issue
→
/plan-issue
→
/start-issue
→
implement
→
/close-issue
| Skill | Command | What it does |
|---|---|---|
| Create Issue | /create-issue | Guided issue creation with title, description, acceptance criteria, squad assignment, and priority. Writes to Linear. |
| Plan Issue | /plan-issue | Break issue into technical plan with file list and approval gate. No code until plan is confirmed. |
| Start Issue | /start-issue | Checkout branch from Linear branchName + mark issue In Progress + post comment. |
| Close Issue | /close-issue | Full quality cycle (review → build → test) + merge + mark Done in Linear + deploy validation if applicable. |
Commercial Skills
Victor — Head of Sales · Lead pipeline, qualification, and deal closing
| Skill | Command | What it does |
|---|---|---|
| Sales Call | /sales-call | 7-part Challenger Sale framework: Warm-up → Reframing → Urgency → Emotional → New Path → Solution → Closing |
| Handle Objection | /handle-objection | Surface-to-real objection mapping, diagnostic flow, non-fit signal detection |
| Follow Up | /follow-up | D+0/D+2/D+7/D+14 follow-up sequence + breakup email template |
| Commercial Debate | /commercial-debate | Multi-perspective roundtable: Sales vs CS vs Finance vs Devil's Advocate |
Customer Success Skills
Leah — Head of CS · Onboarding, retention, and success metrics
| Skill | Command | What it does |
|---|---|---|
| CS Debate | /cs-debate | Roundtable on customer success decisions with multiple perspectives (CSM, Account Manager, Churn Analyst) |
| Log KPI | /log-kpi | Record customer KPI snapshot in data/kpis.md with timestamp |
Dev Skills
Ethan — Tech Lead · Architecture, code review, debugging, and documentation
Code Review Pipeline
Run cheapest → most expensive. Stop when issues are found and fixed.
| Skill | Command | Cost | What it does |
|---|---|---|---|
| Gemini Review | /gemini-review | Free | Gemini 2.5 Flash via OAuth CLI — structural review, P1/P2/P3 classification |
| Codex Review | /codex-review | ~$0.40 | OpenAI Codex CLI — security, logic, performance. Classification: P1 blocks merge · P2 fix before deploy · P3 follow-up |
| Claude Review | /claude-review | Subscription | Claude Opus headless — deep architectural review, final gate before merge |
When to use which:
Trivial change (copy, isolated CSS) → none · New feature →
Trivial change (copy, isolated CSS) → none · New feature →
/codex-review · Logic change → + /claude-review · Critical (auth, billing, migration) → all three + senior approval
Development
| Skill | Command | What it does |
|---|---|---|
| Dev Debate | /dev-debate | Architecture roundtable with multiple tech personas — find blind spots before committing to a design |
| Quick Dev | /quick-dev | Rapid throwaway prototype mode — builds fast, documents nothing, marked as disposable |
| Challenge | /challenge | Independent second opinion on a technical decision without additional context bias |
| Document | /document | Generate documentation after a feature ships — updates code docs and wiki |
Debugging
| Skill | Command | What it does |
|---|---|---|
| Debug | /debug | Pólya 4-phase framework: Understanding → Investigation Plan → Diagnosis → Fix + Retrospective. Triage levels: TRIVIAL / MEDIUM / CRITICAL |
Finance Skills
Clara — CFO · Runway, reporting, and financial health
| Skill | Command | What it does |
|---|---|---|
| Finance Debate | /finance-debate | CFO roundtable on financial decisions — multiple perspectives (conservative, growth, risk) |
| Runway | /runway | Calculate burn rate + months of runway from current squad data |
| Monthly Report | /monthly-report | Generate monthly P&L summary from finance squad data and KPIs |
Infra Skills
Dean — Head of Infra · Servers, monitoring, security, and credential rotation
| Skill | Command | What it does |
|---|---|---|
| Infra Debate | /infra-debate | Infrastructure decision roundtable — cost, reliability, complexity tradeoffs |
| Hardening Check | /hardening-check | Security checklist: open ports, SSH keys, firewall rules, TLS config, env vars, exposed secrets |
| Restart Service | /restart-service | Safe service restart with health check before and after — never blind restart |
| Rotate Credential | /rotate-credential | Guided credential rotation: generate → update services → verify → archive old |
Intelligence Skills
Rex — Head of Intelligence · Competitive analysis, bias audit, war games
| Skill | Command | What it does |
|---|---|---|
| Competitive Analysis | /competitive-analysis | Research competitor, map their positioning, identify gaps and threats relative to your company |
| War Game | /war-game | Simulate competitor moves and model your counter-moves — strategy stress test |
Marketing Skills
Maya — Head of Marketing · Content, brand, campaigns, and distribution
Brand & Strategy
| Skill | Command | What it does |
|---|---|---|
| Marketing Debate | /marketing-debate | Multi-persona roundtable on marketing decisions — Brand, Performance, Content, Devil's Advocate |
| Lookup Brand | /lookup-brand | Answer brand questions from foundation docs: colors, tone, archetypes, anti-patterns |
| Content Ideas | /content-ideas | Generate content ideas aligned with ICP, brand voice, and current trends |
Content Creation
| Skill | Command | What it does |
|---|---|---|
| Write Headline | /write-headline | Generate headline + subheadline + hook variations (4 patterns: curiosity, pain, social proof, direct) |
| Write Caption | /write-caption | Short caption for Instagram/TikTok + LinkedIn variant from the same brief |
| LinkedIn Post | /linkedin-post | Executive LinkedIn format: hook → value bullets → data point → CTA |
| Write Thread | /write-thread | Twitter/Threads thread with reply map and engagement optimization |
| YouTube Script | /youtube-script | Full YouTube script with hook, retention cliffhangers every 2 min, and CTA |
| Email Reactivation | /email-reactivation | SOAP-format reactivation sequence for cold or disengaged leads |
Analysis & Improvement
| Skill | Command | What it does |
|---|---|---|
| Analyze Post | /analyze-post | Diagnose post performance with composite score — hook strength, clarity, CTA, ICP alignment |
| Improve Post | /improve-post | Critical revision of existing post with ready-to-use improved copy suggestions |
+ Add a custom skill
Place your skill file in the squad's skills/ folder, or in the root skills/ for global access.
# Squad-specific skill squads/commercial/skills/my-skill.md # Global skill (all squads) skills/my-skill.md # Skill file template --- name: my-skill description: One-line description --- # /my-skill ## When to use ## Inputs ## Process ## Output format
After adding, register the skill in the squad's CLAUDE.md under ## Skills. See how-to-customize.md for the full guide.