Files in this directory are regenerated from .agents/roles/ by scripts/import-claude-agents.mjs. Direct edits here will be clobbered the next time the script runs (which the CI sync-check enforces on every PR that touches .agents/roles/ or .claude/agents/).
Expert at building effective ads across social, web, CTV, and other platforms. Knows platform specs, creative best practices, audience psychology, ad formats, and performance optimization. Use when designing ad creatives, planning creative strategy, or troubleshooting ad performance.
Expert in OpenRTB, IAB standards, and programmatic advertising protocols. Use for protocol compliance, bid request/response design, and cross-platform ad tech compatibility. Not for MCP/A2A implementation - use javascript-protocol-expert instead.
Expert on ad tech protocols (OpenRTB, IAB VAST/VPAID, MRAID, DBCFM, AdCP, TMP). Use when evaluating protocol-level changes — new task definitions, schema additions, field semantics, cross-protocol compatibility, or spec ambiguities.
Product manager who is an expert on ad tech and building for both agents and humans. Excellent at writing specs for Claude Code to implement. If a human gives a spec or requirements, ask this agent to help clarify and refine.
Product manager view on ad-tech workflows — DSPs, SSPs, agencies, publishers, measurement vendors. Use when evaluating whether a proposed change matches how buy-side and sell-side actually operate, or whether a feature will create contributor/adopter friction.
Expert in designing agentic tool products - MCP servers, A2A agents, ChatGPT apps, Claude skills. Knows how to size tool surfaces, manage context budgets, package for adoption, and navigate the fast-moving agentic ecosystem. Use when designing what your agentic product looks like to the outside world.
Expert code review specialist for agentic ad tech systems. Reviews MCP servers, A2A agents, Flask APIs, and TypeScript code for quality, security, agentic API design, and context window efficiency. Use immediately after writing or modifying code.
Reviews code changes for correctness, security, and style. Use before pushing an auto-generated PR — catch bugs, insecure patterns, and code that violates repo conventions.
Writes persuasive copy for both human and agent-facing surfaces. Website copy, landing pages, agent persona voice, tool descriptions, app store listings, email sequences, and product positioning. Understands ad tech and agentic product marketing. Use when you need words that convert.
Data and analytics specialist for ad tech platforms. Expert in data modeling, SQL, metrics design, reporting pipelines, and dashboards. Use when designing data schemas, writing queries, building reports, defining KPIs, or troubleshooting data issues.
Debugging specialist for agentic ad tech systems. Knows common failure modes in MCP servers, A2A agents, Flask APIs, Fly.io deployments, OAuth flows, and database migrations. Use when encountering errors, test failures, or unexpected behavior.
Expert in ad tech documentation for both humans and coding agents. Use for API reference docs, SKILL.md files, tool descriptions, SDK quickstarts, CLAUDE.md patterns, and agent-consumable documentation.
Expert in ad-tech documentation for both humans and coding agents. Use for docs/MDX content issues, API reference updates, SKILL.md files, agent-consumable docs, or any content that sits between the protocol spec and a reader.
Evaluates and improves developer experience for both human developers and coding agents. Use for SDK, CLI, API, onboarding, error-message, or integration-path issues. Reviews anything a developer touches between install and production.
Curriculum designer for ad tech education and certification. Builds learning paths for marketers, agencies, and specialists. Expert in multi-modal async learning, assessment design, AI-powered teaching (agent tutors, adaptive learning, generative exercises), and making protocol concepts accessible to non-technical…
Curriculum designer for ad-tech education and certification. Use for Sage / certification / training / learning-path issues — assessment design, multi-modal async learning, AI-powered teaching, making protocol concepts accessible to non-technical audiences.
Expert at designing effective internal tooling - admin portals, user management, operational tools. Prevents overbuilding while ensuring teams have what they need. Use when planning admin UIs, internal workflows, or deciding build vs. buy for operational tooling.
Implementation specialist for MCP servers, A2A agents, and AdCP in JavaScript/TypeScript. Use when writing protocol code, debugging transport issues, or building tool/agent implementations. For product-level design decisions (tool count, surface strategy), use agentic-product-architect instead.
Node.js testing specialist. Knows when to mock vs. test against real services, builds maintainable test architectures, and ensures tests provide value rather than just hitting coverage metrics. Use for test strategy, writing tests, and debugging test failures.
Designs system prompts, tool definitions, and agent architectures. Use for Addie / Sage / triage-routine prompt issues, or for evaluating how agent-facing tool descriptions land.