Tekkiiiii
UserOpen-source multi-agent command center for Claude Code. 244 skills, 200+ agents across 14 departments, PD-driven autonomous workflows, SQLite task store, file-based memory. Installs to ~/.claude/ — zero cloud dependencies.
Categories
Indexed Skills (46)
autoplan
Auto-review pipeline — reads the full CEO, design, eng, and DX review skills from disk and runs them sequentially with auto-decisions using 6 decision principles. Surfaces taste decisions (close approaches, borderline scope, codex disagreements) at a final approval gate. One command, fully reviewed plan out. Use when asked to "auto review", "autoplan", "run all reviews", "review this plan automatically", or "make the decisions for me". Proactively suggest when the user has a plan file and wants to run the full review gauntlet without answering 15-30 intermediate questions. (gstack) Voice triggers (speech-to-text aliases): "auto plan", "automatic review".
awesome-design-md
Primary design skill — the default for ALL design tasks across every project. Loaded automatically whenever a design, UI, UX, or visual task is requested. After any design deliverable is produced, this skill runs a mandatory review pass before marking the work complete. Sources 59 curated design system references from the design-refs/ subdirectory, covering the world's best product design (Stripe, Linear, Figma, Airbnb, Notion, etc.).
backend-critique
Senior backend engineer who audits APIs, databases, server logic, auth, file handling, and microservices — acting as a rigorous peer reviewer. Produces a structured critique report with severity ratings (Critical/High/Medium/Low) across 8 dimensions: correctness, API design, database, security, error handling, performance, observability, and maintainability. Use when the user says 'review backend', 'critique this API', 'backend review', 'audit the server code', 'check this database schema', or before shipping backend work. Always read package.json/pyproject.toml/etc., route/controller files, schema files, and relevant context first. Never rewrites code — only flags issues with exact file:line citations and severity. Proactively suggests which issues must be fixed before shipping.
benchmark-models
Cross-model benchmark for gstack skills. Runs the same prompt through Claude, GPT (via Codex CLI), and Gemini side-by-side — compares latency, tokens, cost, and optionally quality via LLM judge. Answers "which model is actually best for this skill?" with data instead of vibes. Separate from /benchmark, which measures web page performance. Use when: "benchmark models", "compare models", "which model is best for X", "cross-model comparison", "model shootout". (gstack) Voice triggers (speech-to-text aliases): "compare models", "model shootout", "which model is best".
benchmark
Performance regression detection using the browse daemon. Establishes baselines for page load times, Core Web Vitals, and resource sizes. Compares before/after on every PR. Tracks performance trends over time. Use when: "performance", "benchmark", "page speed", "lighthouse", "web vitals", "bundle size", "load time". (gstack) Voice triggers (speech-to-text aliases): "speed test", "check performance".
browse
Fast headless browser for QA testing and site dogfooding. Navigate any URL, interact with elements, verify page state, diff before/after actions, take annotated screenshots, check responsive layouts, test forms and uploads, handle dialogs, and assert element states. ~100ms per command. Use when you need to test a feature, verify a deployment, dogfood a user flow, or file a bug with evidence. Use when asked to "open in browser", "test the site", "take a screenshot", or "dogfood this". (gstack)
canary
Post-deploy canary monitoring. Watches the live app for console errors, performance regressions, and page failures using the browse daemon. Takes periodic screenshots, compares against pre-deploy baselines, and alerts on anomalies. Use when: "monitor deploy", "canary", "post-deploy check", "watch production", "verify deploy". (gstack)
career-ops
AI job search command center — evaluates job offers with A-F scoring, generates ATS-optimized CV PDFs, scans multiple job portals in parallel, tracks application pipelines, and analyzes rejection patterns to surface actionable improvements. Routes to specialized sub-agents based on mode: `scan` (portal discovery), `oferta`/`ofertas` (offer scoring and comparison), `pdf` (CV generation), `pipeline` (URL inbox processing), `tracker` (funnel metrics), `batch` (parallel bulk evaluation), `patterns` (rejection pattern detection), `deep` (company research), `apply` (live form assistant), `contacto` (LinkedIn outreach), `training` (course/cert evaluation), `project` (portfolio project idea assessment), `interview-prep` (company-specific STAR+R story bank). Also triggers when pasting a job description or URL — auto-routes to Pipeline Strategist. Ethical gate: scores below 4.0/5 are discouraged; never submits applications on the user's behalf. Built on santifer/career-ops (28K GitHub stars). Ideal for job seekers who
codex
OpenAI Codex CLI wrapper — three modes. Code review: independent diff review via codex review with pass/fail gate. Challenge: adversarial mode that tries to break your code. Consult: ask codex anything with session continuity for follow-ups. The "200 IQ autistic developer" second opinion. Use when asked to "codex review", "codex challenge", "ask codex", "second opinion", or "consult codex". (gstack) Voice triggers (speech-to-text aliases): "code x", "code ex", "get another opinion".
open-gstack-browser
Launch GStack Browser — AI-controlled Chromium with the sidebar extension baked in. Opens a visible browser window where you can watch every action in real time. The sidebar shows a live activity feed and chat. Anti-bot stealth built in. Use when asked to "open gstack browser", "launch browser", "connect chrome", "open chrome", "real browser", "launch chrome", "side panel", or "control my browser". Voice triggers (speech-to-text aliases): "show me the browser".
content-critique
Senior content strategist and copy editor who critiques all written content — marketing copy, technical docs, product UI strings, blog posts, emails, and help text. Produces a structured critique report with severity ratings (Critical/High/Medium/Low) across 6 dimensions: clarity, accuracy, tone & voice, structure, SEO/value, and consistency. Flags AI-slop patterns (filler phrases, throat-clearing openers, business jargon, passive voice, binary contrasts, dramatic fragmentation, rhetorical scaffolding). Use when the user says 'review content', 'critique copy', 'content review', 'audit this text', 'check this doc', 'review this landing page', or before shipping any written content. Never rewrites — flags issues with exact location and severity. Integrates with /stop-slop for AI-pattern detection.
context-restore
Restore working context saved earlier by /context-save. Loads the most recent saved state (across all branches by default) so you can pick up where you left off — even across Conductor workspace handoffs. Use when asked to "resume", "restore context", "where was I", or "pick up where I left off". Pair with /context-save. Formerly /checkpoint resume — renamed because Claude Code treats /checkpoint as a native rewind alias in current environments. (gstack)
context-save
Save working context. Captures git state, decisions made, and remaining work so any future session can pick up without losing a beat. Use when asked to "save progress", "save state", "context save", or "save my work". Pair with /context-restore to resume later. Formerly /checkpoint — renamed because Claude Code treats /checkpoint as a native rewind alias in current environments, which was shadowing this skill. (gstack)
cover-letter-gen
Generate ATS-optimized, company-specific cover letters for career-ops job applications. Paste a job description or provide a company name/URL → tailored cover letter in markdown, ready to attach or submit. Trigger when: the user asks to "generate cover letter", "write cover letter", "cover letter for this job", "/cover-letter", "tailor cover letter", or "company-specific cover letter"; the user pastes a job description; the user names a company and wants a targeted application letter; or the user wants to apply to a role with a customized, ATS-safe cover letter. Key capabilities: extracts company intel from JD (pain points, tech stack, culture signals, repeated keywords); scores achievements from cv.md against JD keywords (3pt direct match, 2pt relevant, 1pt loose, 0pt irrelevant); archetype-based narrative framing (LLMOps, Agentic, Technical AI PM, Solutions Architect, FDE, Transformation); 4-paragraph structure (hook with role+company, proof with top scored achievements, company-specific reasoning, close wi
cso
Chief Security Officer mode. Infrastructure-first security audit: secrets archaeology, dependency supply chain, CI/CD pipeline security, LLM/AI security, skill supply chain scanning, plus OWASP Top 10, STRIDE threat modeling, and active verification. Two modes: daily (zero-noise, 8/10 confidence gate) and comprehensive (monthly deep scan, 2/10 bar). Trend tracking across audit runs. Use when: "security audit", "threat model", "pentest review", "OWASP", "CSO review". (gstack) Voice triggers (speech-to-text aliases): "see-so", "see so", "security review", "security check", "vulnerability scan", "run security".
delegate
Snapshots the full current context — open file, conversation history, project memory files, decisions made, and what's left — then hands the task off to a specialized subagent (preferred) or general-purpose subagent (fallback) that drives to completion autonomously. Use when you need to offload work without losing context, when a task spans files or domains that don't fit a focused specialist, or when you want a subagent to own a task end-to-end while you stay free to work on other things. Also useful for keeping the main context window clean during long multi-step sessions. The subagent receives a structured briefing and works independently — no further prompting required. Check Agency catalog for named specialists first.
careful
Safety guardrails for destructive commands. Warns before rm -rf, DROP TABLE, force-push, git reset --hard, kubectl delete, and similar destructive operations. User can override each warning. Use when touching prod, debugging live systems, or working in a shared environment. Use when asked to "be careful", "safety mode", "prod mode", or "careful mode". (gstack)
browser-harness
Self-healing CDP browser automation connecting LLMs directly to a real Chrome instance via WebSocket. Use when you need full browser freedom with the user's actual sessions, cookies, and extensions — not a headless sandbox. Agents write helper code on-the-fly that persists across runs. Triggers on: "use my real browser", "browser-harness", "automate with my Chrome", "self-healing browser", "domain skill", "real browser session".
cloudflare
Comprehensive Cloudflare platform skill covering Workers, Pages, storage (KV, D1, R2), AI (Workers AI, Vectorize, Agents SDK), networking (Tunnel, Spectrum), security (WAF, DDoS), and infrastructure-as-code (Terraform, Pulumi). Use for any Cloudflare development task. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
content-polish
End-to-end content polishing workflow that runs the humanizer and proofreader in the correct sequence with proper handoff between them. Language-branched: English content uses humanizer + proofreader; Vietnamese content uses humanizer-vi + grammar-checker-vi (with optional translationese-cleaner-vi pre-pass). Use whenever the user wants to take generated content (CVs, posts, articles, emails, cover letters) from draft to ready-to-publish in one step. Triggers: polish this, make this ready, humanize and proofread, clean this up, do the full pass, run the polish workflow. Runs three passes: (1) format-aware humanizer, (2) anti-fragmentation check, (3) format-aware proofreader/grammar check.
cult-ui
Install animated shadcn-compatible components from the Cult UI registry. 75+ components across marketing, onboarding, buttons, cards, interactive elements, visual effects, and typography. Use when adding polished UI components to any shadcn/Next.js project.
dept-resume
Read dept-state.md and spawn a Dept Head with a lean briefing. Department-operations parallel of /pd-resume.
dept-save-state
Write dept-state.md and member-roster.md at end of department session. Department-operations parallel of /save-state.
dept-status
Read-only department status digest. No spawns. Shows active coords, pipelines, open issues, and incoming tasks.
admin-shell-foundation
The shared admin shell scaffold used by all domain skills. Provides the base AdminShell layout, DataTable, StatusBadge, ConfirmDialog, sidebar navigation, and header with user menu. All domain skills (hotel-pms, reservation-booking, restaurant-pos, crm-onboarding) consume this as a shared foundation. Triggers when building any admin dashboard — the layout wrapper that goes around every admin page.
agent-browser
Native Rust headless browser CLI for AI agents. Automates web UIs, scrapes pages, runs QA tests, and performs structured regression testing with health scores and fix loops. Triggers on: "automate browser", "scrape webpage", "QA test this site", "run browser tests", "find bugs on this page", "automate login", "take screenshot", "click this button". Also triggers on: "open this URL", "navigate to", "browser automation", "headless test", "web scraping", "test the web UI". Key capabilities: ~100ms per command, QA workflows with 8-category health scores, diff-aware regression on changed files/routes, multi-session isolation with named profiles, authenticated session persistence, tabs and iframe support, cloud provider integration (Browserless, Browserbase, Browser Use), content boundaries to prevent prompt injection, and command chaining with `&&`. Ideal for QA engineers, developers running regression suites, and agents needing reliable browser automation. Also for: visual bug reports, form filling, cookie sessio
agents-sdk
Build AI agents on Cloudflare Workers using the Agents SDK. Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, chat applications, voice agents, or browser automation. Covers Agent class, state management, callable RPC, Workflows, durable execution, queues, retries, observability, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
animejs
Anime.js adapter patterns for HyperFrames. Use when writing Anime.js animations or timelines inside HyperFrames compositions, registering animations on window.__hfAnime, making Anime.js seek-driven and deterministic, or translating Anime.js examples into render-safe HyperFrames HTML.
auto-researcher
Proactively researches any topic by searching, synthesizing, and presenting well-sourced information — without waiting to be asked. Triggered when the user asks about current events, wants in-depth learning, says "what's the latest on...", "research X for me", "find out about...", "give me a briefing on...", or when the task would benefit from up-to-date external information that might be beyond Claude's training cutoff. Also triggers mid-task proactively when live data, prices, specs, or recent developments would meaningfully improve the answer. Output is structured: a key finding, 3-5 supporting evidence points with sources, any conflicting views, known limitations, and 2-3 further reading links. Source quality is tiered (peer-reviewed primary sources down to anonymous forums), and confidence levels are flagged inline for every major claim. Ideal for competitive research, technical deep-dives, market analysis, and any scenario where fresher information beats remembered knowledge. Also for: keeping a briefin
azure-verified-modules
Azure Verified Modules (AVM) requirements and best practices for developing certified Azure Terraform modules. Use when creating or reviewing Azure modules that need AVM certification.
backend
Designs, builds, and reviews backend systems: APIs, databases, server-side logic, authentication, file handling, webhooks, and microservices. Triggers when the user asks to build an API, design a database schema, write server-side code, set up authentication, handle file uploads, build webhooks, design microservices, optimize queries, or work with Node.js, Python, Go, Java, or any server-side technology. Also triggers proactively when reviewing backend code for performance, scalability, or correctness issues — including N+1 queries, missing indexes, connection pool exhaustion, and missing pagination. Key capabilities: layered architecture patterns (routes/controllers/services/repositories), REST design with correct HTTP status codes, UUID-based public IDs with created_at/updated_at timestamps, JWT auth with short-lived access tokens + httpOnly refresh cookies, bcrypt password hashing at cost factor 12+, centralized error handling, structured JSON logging, parameterized queries only, and background job pattern
brandkit
Premium brand-kit image generation skill for creating high-end brand-guidelines boards, logo systems, identity decks, and visual-world presentations. Trained for minimalist, cinematic, editorial, dark-tech, luxury, cultural, security, gaming, developer-tool, and consumer-app brand systems. Optimized for intentional logo concepting, refined composition, sparse typography, strong symbolic meaning, premium mockups, art-directed imagery, and flexible grid layouts.
browser-domain-skills
Shared domain knowledge convention for all browser automation tools. Stores per-host notes (login flows, selector quirks, anti-bot workarounds, form structures) that any browser skill can read before navigating and write after learning. Self-healing pattern: observe → capture → reuse across sessions. Not a tool — a shared memory protocol.
cc-loop
Iterative quality loop for any content deliverable. Runs fixer → polish → N parallel critiques → score → loop until pass criteria are met. Default pass criteria: avg score ≥ 80 AND min score ≥ 70 across all critics. Max 3 rounds (configurable). Critic set is domain-configurable: design+content+marketing+pedagogy for decks/presentations; content+marketing+SEO for blog posts; content+marketing for emails; custom list accepted. Each round writes an audit log entry so you can track how the deliverable evolved. Trigger when: user says "run X through the loop", "polish and critique", "iterate to quality bar", "/cc-loop", or any time a deliverable needs systematic multi-axis review with pass criteria. Also trigger when a single critique came back with issues and the user wants a structured fix-and-verify cycle rather than ad-hoc edits. Key capabilities: snapshot-grounded design critique via Playwright at 1920×1080, format-aware humanizer + proofreader polish pass, configurable scoring threshold and max rounds, refra
cli-anything
Turns any GUI-only software into an agent-usable CLI/REPL harness backed by the software's real engine (Blender, GIMP, LibreOffice, etc.).
cloudflare-email-service
Send and receive transactional emails with Cloudflare Email Service (Email Sending + Email Routing). Use when building email sending (Workers binding or REST API), email routing, Agents SDK email handling, or integrating email into any app — Workers, Node.js, Python, Go, etc. Also use for email deliverability, SPF/DKIM/DMARC, wrangler email setup, MCP email tools, or when a coding agent needs to send emails. Even for simple requests like "add email to my Worker" — this skill has critical config details.
content-strategy
Build editorial calendars, define content pillars, plan campaigns across TOFU/MOFU/BOFU funnel stages, audit and repurpose content across channels, and track performance by channel and pillar. Trigger when: building or maintaining an editorial calendar; defining content pillars or themes for a niche or ICP; planning a content campaign (blog, video, podcast, social); auditing or repurposing existing long-form content; developing a content distribution or SEO strategy; or analyzing content performance.
context-mode
Context-mode is an MCP server plugin that reduces context window usage by ~98% through sandboxed tool routing. MUST be applied whenever Claude is running inside a context-mode-enabled session — it is not optional. Trigger when: context-mode MCP server is present in the session; running Bash commands that produce more than 20 lines of output; reading files for analysis; fetching web pages and searching them; batch execution of multiple queries; or when you need to prevent raw tool outputs from flooding the context window. Key capabilities: routes Bash to ctx_execute / ctx_batch_execute (captures output, prevents bloat); routes Read to ctx_execute_file (raw content never enters context); routes WebFetch to ctx_fetch_and_index then ctx_search (fetch once, query indexed content on demand); enforces strict output rules (under 500 words, write artifacts to files, return only path + 1-line description). Ideal for: working in long-running sessions with large codebases; running multi-step debugging workflows that prod
coord-respawn-self
Coord context-aware self-respawn. Invoked by a Coord agent when context window reaches 80% threshold mid-L3. Saves mid-L3 state, writes a continuation manifest, notifies the spawner (PD or Dept Head depending on caller role), and stops. The spawner then resumes the L3 with a fresh Coord. Enforces max 3 respawns per Coord per 24h. Use at clean task boundaries only — never mid-Executor-spawn or mid-ACK/NACK cycle.
copywriting
Write compelling, conversion-focused copy for any medium. Trigger when: writing marketing copy, landing page text, email campaigns, ad copy, product descriptions, social media posts, headlines, taglines, or sales pages; improving existing copy for clarity, persuasiveness, or brand voice; increasing conversion rates; developing or auditing brand voice; or any persuasive writing task. Before writing, clarify: audience (who, fears, wants), goal (click, purchase, sign up, share), channel (email, landing page, ad, social), tone (formal, casual, urgent, inspirational), and the one key message. Key capabilities: AIDA, PAS, and Before/After/Bridge frameworks; brand voice capture system (3 words the brand IS, 3 words it is NOT, who speaks, what they never say); headline writing rules (benefit-led, numbers, questions, urgency without fake scarcity); email copy structure (subject → preview → opening → body → single CTA); copy audit checklist (5-second benefit test, one-person test, single CTA check, "so what?" on every
crm-onboarding
Build a lead capture + email nurture CRM pipeline — public onboarding form, CRM state machine (NEW→QUALIFIED→PROPOSAL→WON/LOST), 5-email drip sequence via Resend, Slack webhook handoff on qualification, and kanban dashboard. Triggers when user asks to build a CRM, lead capture, onboarding flow, lead nurture, or lead pipeline.
css-animations
CSS animation adapter patterns for HyperFrames. Use when authoring CSS keyframes, animation-delay based timing, animation-fill-mode, animation-play-state, or CSS-only motion that HyperFrames must seek deterministically during preview and rendering.
deck-narrative
The ARCHITECTURE layer beneath every deck skill — decides WHAT goes on slides and in WHAT ORDER, before any visual skill decides what it looks like. Covers action titles vs topic titles (the verb test), one-message-per-slide, Minto top-down sequencing (conclusion on slide 2, never a surprise ending), pacing by meeting length, horizontal logic (titles alone read as a persuasive essay) and vertical logic (body proves title), table-vs-chart selection, and decision/approval-slide construction (Challenge>Options>Decision, Goal>Obstacle>Strategy>Action, Insight>Action>Outcome). Every rule is written as an auditable pass/fail check, not a style opinion. Ships `scripts/audit_deck.py` — a runnable Python auditor that reads a .pptx and mechanically reports per-slide font size, title word/line counts, a verb-test heuristic, table vs chart counts, reader-only-content (process- meta marker) review candidates, a deck-wide font-family inventory with a hard system-font (Calibri/Arial/etc.) fail flag, and a title-sequence dum
compile
Load the marketing assessment pipeline lessons, pitfalls, and quality checklist before generating or reviewing a client deck. Invoke with /compile before any pipeline run to ensure all 24 learned pitfalls are loaded into context. Also use after a pipeline run to verify output against the quality checklist.
content-experimentation-best-practices
Content experimentation and A/B testing guidance covering experiment design, hypotheses, metrics, sample size, statistical foundations, CMS-managed variants, and common analysis pitfalls. Use this skill when planning experiments, setting up variants, choosing success metrics, interpreting statistical results, or building experimentation workflows in a CMS or frontend stack.
content-modeling-best-practices
Structured content modeling guidance for schema design, content architecture, content reuse, references versus embedded objects, separation of concerns, and taxonomies across Sanity and other headless CMSes. Use this skill when designing or refactoring content types, deciding field shapes, debating reusable versus nested content, planning omnichannel content models, or reviewing whether a schema is too page-shaped or presentation-driven.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.