Wonderful

Deploy an AI agent or agent harness to a Nebius Managed Kubernetes cluster using the KubeClaw production toolkit pattern (Helm chart, observability stack, secrets, autoscaling, Token Factory wiring). Produces a values.yaml, ingress config, secret-management plan, observability wiring (Prometheus + Loki + Grafana), readiness checks, autoscaling policy, and a verified rollout plan. Use whenever a user wants to deploy an agent to Kubernetes, ship an OpenClaw or NemoClaw stack to Nebius, productionize a prototype agent, asks how to host an agent in a cluster, wants a Helm chart for an LLM service, or needs autoscaling/observability for an LLM workload.

Resolve, verify, and audit an agent's identity using the AID (Agent Identity & Discovery) v2 protocol. Performs DNS-based discovery, signature verification, capability negotiation, and produces a structured identity report with trust signals. Use whenever a user wants to verify which agent is calling their service, look up an agent's capabilities and provenance from its handle, audit agent-to-agent traffic, asks 'who is this agent', 'is this AID valid', 'verify this agent signature', 'what can agent X do', or is building a service that needs to authenticate non-human callers. Trigger for agentic commerce, x402 micropayments, agent-to-agent (A2A) flows, and any system gating actions on caller identity.

Audit a business, team, role, workflow, or tool stack to identify practical AI assistant and AI agent opportunities. Use when a user wants to know what to automate, where agents could save time or increase revenue, how to prioritize AI implementation, or what first agent/skill to build.

Audit an AI agent or LLM-powered system against the agentic-engineering discipline. Scores Harness, Context, Tool Design, Skills, Memory, Evals, Model Selection, Failure Modes, and Adversarial Collaborator patterns, then prescribes three highest-ROI fixes. Use when a user wants to review an agent architecture, diagnose why an agent is unreliable or expensive, prepare a production launch, critique an LLM system design doc, or ask whether an agent is production-ready.

Comprehensive working reference for agentic engineering: the discipline of building reliable autonomous coding agents and AI-native software systems. Invoke when the user asks about designing or evaluating agent systems, harness and context engineering, tool design for agents, agent memory, evals, model selection for agentic work, the named architectural patterns (isolation, orchestration, feedback loops, failure recovery, multi-agent coordination), choosing between coding agents (Claude Code, Cursor, Devin, OpenHands, the OpenClaw ecosystem, etc.), avoiding agentic technical debt, or any question of the form 'how do I build, run, or scale agents reliably.' Use even when the user does not say 'agentic engineering' but is clearly building or operating autonomous agents.

Turn recurring tasks, SOPs, workflows, prompts, business processes, or expert know-how into reusable AI skills with trigger descriptions, step-by-step instructions, constraints, examples, and quality checks. Use when a user wants to create a Codex skill, Claude skill, agent skill, prompt pack, or reusable AI workflow.

Build a complete ambassador program — tiers, benefits, application flow, onboarding, and tracking — for developer communities, DevRel programs, or AI/web3 ecosystems. Use for ambassador program, champions program, advocate tier, referral program for developers, or builder program structure.

Audit a codebase, environment, or config for exposed API keys and secrets, and recommend safe handling, vaulting, and rotation. Scans for hardcoded keys, secrets in committed files, keys in client-side bundles, and weak storage patterns, then produces a prioritized findings report and a remediation plan. Defensive use only. Use whenever the user wants to check for leaked credentials, secure their AI/API keys, asks 'do I have any exposed keys', 'is this safe to commit', 'how should I store my API keys', or is prepping a repo to go public. Trigger before open-sourcing a repo or after suspecting a leak. Never use to locate or exploit others' credentials.

Track how often a brand, product, or person shows up in answers from major LLM assistants (ChatGPT, Gemini, Claude, Perplexity, Copilot) for a defined set of queries, then produce a daily or weekly visibility report card with rank, share-of-voice, sentiment, and competitor comparison. Use whenever a user wants to monitor LLM brand visibility, measure AI search presence, track AEO (AI Engine Optimization), compare their share-of-voice vs competitors in LLM answers, asks 'how often does ChatGPT mention us', 'are we showing up in AI search', 'track our brand across LLMs', or wants the AI-era equivalent of SEO rank tracking.

Turn messy descriptions of business operations into clear, reusable workflows. Use when a user shares a rough process, daily task, voice note, transcript, SOP fragment, back-office routine, client delivery process, or wants to identify what an AI assistant or agent should do step by step.

Turn client briefs, kickoff notes, proposals, discovery calls, scope notes, or messy project context into delivery plans, checklists, handoff docs, status updates, and implementation briefs. Use for service businesses, consultants, agencies, and operators managing client work.

Benchmark coding agents against open-source models with CodingAgentBench to find which agent plus which model actually codes best for a given task profile. Takes a task suite (or helps define one) and a set of agent/model combinations, runs them, scores results on pass rate, quality, cost, and latency, and emits a ranked leaderboard with a recommendation. Use whenever the user wants to compare coding agents or models, asks 'which model should I use for coding', 'benchmark these agents', 'which open model is best with Claude Code / OpenClaw', 'is model X good enough for this task', or wants evidence-based model selection rather than a guess. Trigger for any agent-times-model evaluation.

Generate a complete community launch kit — Discord/Slack/Telegram server structure, welcome message, channel list, rules, onboarding flow, and 30-day content calendar. Use for set up my community, structure my Discord, community launch plan, or 30-day content calendar.

Plan and run the operations of a community or events space, a hacker house, coworking floor, community house, or events venue. Produces event programming calendars, member and guest communications, house operating procedures, space-and-resource scheduling, and budget/revenue framing. Use whenever the user is operating or launching a physical community space and needs help with programming, member comms, house rules, event logistics, or making the space financially work, says 'help me program our space', 'run our community house', 'event calendar for our venue', or 'operating plan for our coworking floor'. Trigger for both the programming side (what happens in the space) and the operations side (how it runs).

Repurpose source material into a practical content engine for founders, consultants, creators, agencies, and service businesses. Use when a user provides an idea, transcript, essay, podcast notes, sales call, workshop, newsletter draft, or wants posts, hooks, angles, scripts, newsletters, or a repeatable publishing system.

Guide for using ConTree sandboxed container execution with Git-like branching. Use this skill whenever the user mentions ConTree, contree, or contree-mcp, OR wants to run code in isolated sandboxes/containers with branching capabilities, OR the ConTree MCP tools are available (contree_run, contree_rsync, contree_import_image, contree_list_images, contree_upload, contree_download, contree_list_files, contree_read_file, contree_set_tag, contree_get_operation, contree_wait_operations). Also use when the user wants to safely execute untrusted code, branch execution states, explore multiple solution paths in parallel, or do tree-search style coding. Even if they just say 'run this in a sandbox' or 'try multiple approaches in parallel containers' — this skill applies.

Create or update a Coral source spec YAML for a custom HTTP API or local dataset. Use when authoring a standalone source for `coral source add --file`, or when adapting that spec into a Coral repo source under `sources/core` or `sources/community`.

Review new or updated Coral source manifests and source PRs for content, style, product fit, query ergonomics, documentation quality, and consistency with existing Coral sources. Use when Codex is asked to review a sources/core/name or sources/community/name source directory, a manifest.yaml, or a GitHub PR that adds or changes a Coral source.

Upgrade a plain slide deck or a rough outline into an interactive, web-based slideshow spec themed to a brand. Takes deck content (bullets, an outline, exported slides, or notes) and produces a slide-by-slide build spec: narrative arc, per-slide layout and visual treatment, interactive elements, speaker notes, and brand-themed styling. Use whenever the user wants to improve, redesign, or modernize a presentation, turn slides into a web slideshow, says 'upgrade my deck', 'make these slides interactive', 'redesign this presentation', or wants a flat deck turned into something dynamic. Trigger even if they just paste slide content and ask to make it better.

Turn a raw product demo or rough description into a punchy, high-conversion pitch. Produces title options, a one-line hook, a 60-second spoken demo script, thumbnail and caption copy, and a tight written blurb, tuned for a demo feed, pitch event, or social clip. Use whenever the user has a product, demo, or project and wants to present it compellingly, says 'help me pitch this', 'make my demo punchy', 'write a demo script', 'thumbnail and caption for this', or is prepping for a demo day, launch, or pitch feed. Trigger even if they just describe what they built and ask how to show it off.

Write trust-first email and newsletter copy aimed at a developer and builder audience. Produces subject line variants, preview text, body copy, segment suggestions, and a deliverability checklist, in a voice that respects technical readers and avoids marketing fluff. Use this whenever the user is writing an email, newsletter, announcement, launch note, or campaign aimed at developers, builders, or a technical community, even if they do not say 'developer email.' Trigger on phrases like 'write a newsletter', 'email to announce', 'draft a campaign', 'send an update to our list', or any request to produce email copy for a technical or builder audience.

Generate a scoped statement-of-work for a fractional Developer Relations engagement. Produces objectives mapped to a real activation funnel, mechanisms drawn from a fixed menu, three pricing tiers, kill-switches, and a proof section. Use when a user wants to draft a DevRel SOW, scope a fractional DevRel engagement, price a developer event or community contract, define non-vanity DevRel KPIs, or evaluate an inbound DevRel proposal.

Turn a single source asset (a talk, demo, repo, launch, blog post, or transcript) into a full set of channel-ready content: an X/Twitter thread, a LinkedIn post, a newsletter blurb, and a short blog post. Preserves technical accuracy while adapting tone and length per channel. Use this whenever the user wants to repurpose, atomize, or distribute one piece of content across multiple platforms, even if they do not say 'repurpose.' Trigger on phrases like 'turn this talk into', 'make social posts from', 'I gave a demo and want to write about it', 'repurpose this', 'content from our launch', or any request to spin one input into multiple formats for a developer audience.

Generate a 3–5 email developer relations sequence — onboarding, product launch, event promotion, or nurture — with subject lines, preview text, and full body copy for Beehiiv, Mailchimp, or HubSpot. Use for write email sequence, drip campaign, onboarding emails, or developer newsletter.

Produce an outside-in developer relations strategy and a leave-behind for a company or product. Takes a product (and optionally its docs, site, or positioning) and produces a DevRel strategy: developer ICP, the adoption funnel and where it leaks, content and community plays, channel strategy, a 90-day plan, and success metrics, plus a polished leave-behind a hiring manager or founder can read. Use whenever the user wants a DevRel strategy, plan, or pitch, says 'build a DevRel strategy for', 'how should we do developer relations', 'pitch myself for this DevRel role', or is prepping a leave-behind or interview deliverable. Trigger for both inbound (our own product) and outbound (pitching a company) framing.

Generate a complete DevRel strategy document with positioning, OKRs, 90-day plan, channel strategy, and budget framework. Use for DevRel strategy, DevRel proposal, fractional DevRel pitch, developer relations roadmap, DevRel one-pager, or DevRel 90-day plan.

Generate a Luma-ready event description, email blast copy, and LinkedIn + X/Twitter social copy from minimal inputs. Use for write event copy, create a Luma event, draft event brief, help me market my event, or generate event description.

Full lead-analysis pipeline for event/hackathon registration CSVs: verify every attendee against live web sources, score them per sponsor product, rank by social influence (X/GitHub followers), flag competitors/investors/sponsor-staff, and ship a sortable spreadsheet and/or password-gated web dashboard, plus curated Top-N builder and judge lists. Use this whenever the user has an attendee/registration/guest CSV (Luma, Eventbrite, etc.) and wants leads reviewed, screened, enriched, categorized, ranked, or turned into a dashboard; whenever they ask 'who are the top leads/builders/judges at this event'; for follow-up requests on an existing event analysis; and when they want to productize this (pull from the Luma API, scrape event calendars, or run autonomously). Trigger even if they just say 'review the attached attendee list for our sponsors'.

Generate a complete, ready-to-publish event page for a developer or tech community event from a rough idea. Produces title options, a hook, agenda, host bio, logistics block, a clear CTA, and a registration reminder sequence, formatted for Luma or any event platform. Use this whenever the user is creating, planning, or writing copy for an event, meetup, workshop, mixer, summit, launch party, or community gathering, even if they only describe the idea loosely and do not explicitly say 'event page.' Trigger on phrases like 'I'm hosting', 'we're running an event', 'help me write a Luma page', 'create an event description', or any request to turn an event concept into published copy.

Decide which events are worth attending and who to pitch there, or turn event attendees into a scored, deduped, CRM-ready lead list. Use when a user is choosing conferences, meetups, hackathons, or summits to attend, asks whether an event is worth it, wants target people or companies to meet, provides a Luma/Lu.ma URL or attendee CSV, wants to score event leads, dedupe attendees, diff against a CRM, or plan event follow-up.

Produce the post-event recap deliverable that goes to sponsors, stakeholders, and the community after a hackathon, builder event, or developer summit. Covers attendance numbers, demographic mix, demo highlights, sponsor-specific outcomes (signups, integrations shipped, qualified leads), social and press footprint, photos and videos, and a renewal ask. Use whenever a user has just run an event and needs to report results, says 'write our event recap', 'sponsor report from the hackathon', 'post-event summary for our team', 'recap email to the community', or needs the after-action document that turns one event into the next sponsor's contract. Trigger from event end through T+14 days.

Diagnose what stage an AI-native or developer-tools founder is actually in, then return three things to do this week and three things to stop doing. Use when a founder asks what to focus on, whether they have product-market fit, whether to hire or raise or pivot, why they feel stuck, or wants a brutal external read on their company stage. Optimized for early-stage AI and devtools startups.

Generate a complete operational plan for a developer hackathon, AI build day, code sprint, builder summit, or sponsor-backed competitive event. Produces event summary, run of show, judging rubric, prize and sponsor tiers, judge brief, participant comms, logistics checklist, lead capture guardrails, budget worksheet, and risk register. Use when the user is planning, organizing, pitching, or improving a hackathon or competitive builder event.

Run a hackathon while it is happening — the in-event operational layer covering judge scheduling, team formation, demo capture, leaderboard, incident triage, and real-time announcements. Distinct from pre-event planning (which is covered by hackathon-in-a-box and hackathon-rubric-builder) and post-event reporting (covered by event-recap-report). Use whenever a user is actively running a hackathon and needs help during the event, says 'we're at the event and X', 'schedule the judges', 'how do we handle the demo queue', 'leaderboard for the event', 'announcement to attendees mid-event', 'team needs a substitute member', or any live-event coordination request. Trigger from T-24 hours through the demo session.

Build a complete hackathon judging rubric, per-category scoring guide, and judge briefing doc from minimal inputs. Works for AI, web3, and general developer hackathons. Use for build a rubric, judging criteria, hackathon scoring, judge briefing, or evaluation criteria.

Pick an LLM inference stack — provider, model, and hardware — given throughput, latency, cost, sovereignty, and compliance constraints. Compares hyperscaler APIs (OpenAI, Anthropic, Google), dedicated inference clouds (Nebius, Together, Fireworks, Groq, Cerebras), and self-hosted options (vLLM, SGLang, tinygrad). Use when a user asks where to run a model, how to reduce LLM cost, whether to self-host versus use an API, how to plan GPU sovereignty, or to compare inference providers.

Generate a developer-ready quickstart and starter README for integrating a tool, API, SDK, or platform, in the proven pattern of a starter repo. Produces a structured README with prerequisites, install steps, a minimal working example, common gotchas, and next steps, plus a suggested repo structure. Use this whenever the user wants to document an integration, write a quickstart, create a starter repo, onboard developers to a tool, or lower the time-to-first-success for an API or SDK, even if they do not say 'quickstart.' Trigger on phrases like 'write a quickstart for', 'starter repo for', 'getting started guide', 'how do devs integrate our', 'README for our SDK', or any request to make a developer tool easy to adopt fast.

Scaffold a security-first Model Context Protocol (MCP) server for any backing service, with tool typing, auth, rate limiting, audit logging, and a smoke-test suite. Produces a complete Python or TypeScript MCP server project with auth middleware, per-tool rate limits, typed tool schemas, structured logs, and a Claude Desktop / Cursor / Codex client config that wires it up. Use whenever a user wants to expose a service to AI agents as an MCP server, build a tool layer for an API or database, asks how to add MCP support to their product, wants to convert an internal API into agent-callable tools, or needs the security/observability patterns for an MCP server destined for production.

Convert meeting notes, transcripts, call recordings, sales calls, client calls, team syncs, or rough discussion notes into decisions, action items, owners, deadlines, risks, follow-up messages, and project momentum. Use when a user wants useful next steps from a meeting.

Survey upcoming events across a city or region within a date window and rank them by ICP density — which events are worth attending or sponsoring for a given target audience. Produces a ranked event calendar with attendance estimates, audience-fit scores, sponsor presence, cost-to-attend, and ROI projections. Distinct from event-radar (which extracts attendees from one event) and event-leads (which scores attendees of a single event). Use whenever a user wants to plan which events to attend or sponsor over the next weeks, asks 'what AI events should I go to next month', 'rank Devcon side events for us', 'which SF builder events are worth sponsoring', 'plan my event calendar for Q3', or wants to triage many candidate events down to a focused list.