scottconverse

Doc Writer for ProductTeam. Reads source code and produces all documentation: README.md, full docs, plain text, landing page, PDF, changelog. Never fabricates features — documents only what exists in the code.

Design Evaluator for the product development pipeline. Grades visual artifacts (landing pages, documentation, CLI output, PDFs) against four criteria: Coherence, Originality, Craft, and Functionality. Never fixes — only grades and reports.

Planner agent for the product development pipeline. Takes a PRD or feature request and produces a structured sprint contract with specific deliverables and testable acceptance criteria. The planner never writes code — only specs.

PRD Writer for ProductTeam. Takes a product concept from a PM or product owner, asks clarifying questions, researches competitors, and produces a structured PRD that the Planner can consume. The entry point of the ProductTeam pipeline.

Builder agent for the product development pipeline. Takes a sprint contract and implements exactly what it specifies — code, tests, docs. The Builder never declares work 'done' — only 'ready for review.' Only the Evaluator can declare done.

Evaluator agent. Verifies Builder output against sprint contract. Quality level (standard/thorough/strict) controls evaluation depth and cost.

Orchestrator for ProductTeam. Manages the full product development pipeline: PRD Writer -> Planner -> Builder -> Evaluator -> Doc Writer. Routes work to the right agents, manages build-evaluate loops (max 3), enforces approval gates, and writes session handoff artifacts. The brain of ProductTeam.

UI Builder for ProductTeam. Specialized Builder for frontend and visual work: web UIs, landing pages, dashboards. Follows a default design system with dark theme, WCAG AA accessibility, and responsive-first approach. Graded by the Design Evaluator.

Use when the user wants to produce a complete CIO artifact bundle — IT strategy, technology roadmap, vendor portfolio, digital transformation plan, IT governance framework, enterprise architecture, budget allocation model, workforce development plan, and IT risk appetite statement — for an organization. Triggers when the user says "create an IT strategy", "technology roadmap", "IT governance", "digital transformation", "CIO agent", "/cio-agent", or describes a CIO or IT leadership artifact task. Invokes the AgentSuite CIO agent via MCP.

Use when the user wants to produce a complete trust and risk artifact bundle — threat model, risk register, control framework, incident response plan, compliance matrix, vendor risk assessment, security policy, audit readiness report, and residual risk acceptance — for a product, service, or organization. Triggers when the user says "create a threat model", "risk assessment", "security policy", "compliance matrix", "trust risk agent", "/trust-risk-agent", or describes a trust or risk artifact task. Invokes the AgentSuite Trust/Risk agent via MCP.

Use when the user wants to produce a complete marketing specification bundle — campaign brief, target audience profile, messaging framework, content calendar, channel strategy, SEO keyword plan, competitive positioning, launch plan, and measurement framework — for a product, service, or brand initiative. Triggers when the user says "create a campaign brief", "write a messaging framework", "marketing strategy", "content calendar", "channel strategy", "marketing agent", "/marketing-agent", or describes a marketing artifact task. Invokes the AgentSuite Marketing agent via MCP.

Use when the user wants to produce a complete engineering specification bundle — architecture decision record, system design, API spec, data model, security review, deployment plan, runbook, tech-debt register, and performance requirements — for a system, service, or technical initiative. Triggers when the user says "create an ADR", "write a system design", "architecture decision record", "API spec", "engineering spec", "engineering agent", "/engineering-agent", or describes an engineering artifact task. Invokes the AgentSuite Engineering agent via MCP.

Use when the user wants to produce a complete product specification bundle — PRD, user story map, feature prioritization, success metrics, competitive analysis, user personas, acceptance criteria, product roadmap, and risk register — for a product, feature, or initiative. Triggers when the user says "create a PRD", "write a product spec", "product roadmap", "feature requirements", "product agent", "/product-agent", or describes a product management artifact task. Invokes the AgentSuite Product agent via MCP.

Use when the user wants to produce a complete design specification bundle — visual direction, design brief, mood board, brand rules, image generation prompts, accessibility audit, and asset acceptance checklist — for a campaign, product, or brand. Triggers when the user says "design a campaign", "create design specs", "design agent", "/design-agent", or describes a visual direction / brand identity / asset brief task. Invokes the AgentSuite Design agent via MCP.

Use when the user wants to build a reusable creative-ops system for a product or company — brand system, voice guide, audience map, brief library, prompt library. Triggers when the user says "build a brand system", "set up creative ops", "founder agent", "/founder-agent", or describes a launch/positioning/brand-foundation task. Invokes the AgentSuite Founder agent via MCP.

Use when evaluating one worker result against its task contract and writing a structured judgment.

Use when decomposing a broad product or engineering goal into isolated Codex worktree tasks with strict contracts.

Use when merging accepted worker outputs into a coherent final change set with minimal new scope.

Use when producing a final operator release checklist covering tests, docs, risks, and unresolved issues.

Use when executing a single task contract inside a Codex worktree with minimum viable context.

Use context-mode tools (ctx_execute, ctx_execute_file) instead of Bash/cat when processing large outputs. Triggers: "analyze logs", "summarize output", "process data", "parse JSON", "filter results", "extract errors", "check build output", "analyze dependencies", "process API response", "large file analysis", "page snapshot", "browser snapshot", "DOM structure", "inspect page", "accessibility tree", "Playwright snapshot", "run tests", "test output", "coverage report", "git log", "recent commits", "diff between branches", "list containers", "pod status", "disk usage", "fetch docs", "API reference", "index documentation", "call API", "check response", "query results", "find TODOs", "count lines", "codebase statistics", "security audit", "outdated packages", "dependency tree", "cloud resources", "CI/CD output". Also triggers on ANY MCP tool output that may exceed 20 lines. Subagent routing is handled automatically via PreToolUse hook.

Show how much context window context-mode saved this session. Displays token consumption, context savings ratio, and per-tool breakdown. Read-only — shows stats only, no reset capability. To wipe the knowledge base entirely, use ctx_purge instead. Trigger: /context-mode:ctx-stats

Run context-mode diagnostics. Checks runtimes, hooks, FTS5, plugin registration, npm and marketplace versions. Trigger: /context-mode:ctx-doctor

Purge the context-mode knowledge base. Permanently deletes all indexed content and resets session stats. This is destructive and cannot be undone. Trigger: /context-mode:ctx-purge

Snapshot session state to .claude/handoff_current.md (plus a raw-dump backup at .claude/handoff_backups/handoff_raw_<session_id>.md that the Stop hook has been appending to all session, when Stop hooks fire) and tell the user (loudly, with -*-*- borders) to start a new session. Use at clean boundaries (commit lands, track wraps), when the user signals context pressure ("getting long", "meter is full"), or whenever the user invokes /handoff. Blocks until the user actually starts a new session — do not start new work after invoking.

Run a visitor-grade audit of a project's public-facing surfaces — verify pages, docs, READMEs, release pages, and announcements the way a first-time human visitor experiences them, not the way a developer diffs them. Reads every surface fully as rendered, follows EVERY link (counted, with per-link status), checks published checksums, and flags stale claims, jargon leaks, invisible links, mojibake, and misleading framing. Use whenever the user says "visitor audit", "audit the public surfaces", "click through everything", "is this actually done", or "/visitor-audit" — and proactively before announcing any page, doc, site, or release as done, and again after any deploy, even if the user doesn't say "audit". CI green is not a substitute for this check.

Use when Scott says "check the repo for latest test directive", "check the repo for test directive", "run the latest CivicCast test directive", or asks this tester machine to run the CivicCast proof from the tester branch.