Planning & Project Management for Claude Code

Synthesizes classified planning docs into a single consolidated context. Applies precedence rules, detects cross-ref cycles, enforces LOCKED-vs-LOCKED hard-blocks, and writes INGEST-CONFLICTS.md with three buckets (auto-resolved, competing-variants, unresolved-blockers). Spawned by /gsd:ingest-docs.

Designs a structured evaluation strategy for an AI phase. Identifies critical failure modes, selects eval dimensions with rubrics, recommends tooling, and specifies the reference dataset. Writes the Evaluation Strategy, Guardrails, and Production Monitoring sections of AI-SPEC.md. Spawned by /gsd:ai-integration-phase orchestrator.

Analyzes codebase for existing patterns and produces PATTERNS.md mapping new files to closest analogs. Read-only codebase analysis spawned by /gsd:plan-phase orchestrator before planning.

Researches how to implement a phase before planning. Produces RESEARCH.md consumed by gsd-planner. Spawned by /gsd:plan-phase orchestrator.

Verifies plans will achieve phase goal before execution. Goal-backward analysis of plan quality. Spawned by /gsd:plan-phase orchestrator.

Creates executable phase plans with task breakdown, dependency analysis, and goal-backward verification. Spawned by /gsd:plan-phase orchestrator.

Researches domain ecosystem before roadmap creation. Produces files in .planning/research/ consumed during roadmap creation. Spawned by /gsd:new-project or /gsd:new-milestone orchestrators.

Synthesizes research outputs from parallel researcher agents into SUMMARY.md. Spawned by /gsd:new-project after 4 researcher agents complete.

Creates project roadmaps with phase breakdown, requirement mapping, success criteria derivation, and coverage validation. Spawned by /gsd:new-project orchestrator.

Critic and devil's advocate specialist. Challenges strategy before commitment -- finds the strongest counterargument, names what was missed, probes for hidden assumptions. Read-only. Use before any significant decision. Part of the operator-kit five-agent starter kit.

Bias-isolated agent that reads a customer discovery interview transcript (or recollection) and produces a structured interview analysis file — extracted statements with hypothesis backlinks, plus technique feedback. Dispatched by the interviews skill. Does not evaluate hypothesis state or write to hypothesis files.

Bias-isolated assessment agent that evaluates a startup project's current state against lean startup methodology and recommends plan updates. Dispatched by the whats-next skill — not invoked directly by the founder. Returns structured recommendations as text; does not write files or interact with the user.

The Boss of the life team. Seasoned coach who has sat with many decisions, many doubts, and many regrets. Listens without judgment, then gives a grounded call.

The Rookie of the life team. Close friend energy. Asks the questions the user has been avoiding. Not contesting, helping them hear themselves think.

The Watcher of the life team. Philosophical, lateral-thinking friend. Pulls in angles the user did not consider, questions the frame, sometimes reframes the entire decision.

Tier 3 (deep) Ralph Review. Dispatched by the /ralph-review command after Tier 2 passes. Runs thorough architectural review — fit, standards, governance drift, null propagation, config wiring, factual-claims audit. Never writes code.

Agent that analyzes projects to create WORK (unit of work) and decompose sub-TASKs. Must be used for requests like "plan this", "decompose TASKs", "build XXX", "add XXX feature". Reads CLAUDE.md, README, and source code to create WORK and derive sub-TASKs.

Delegates planning work to a Codex planner agent (read-only, no worktree). Use when you want a second pass on an implementation plan, or when plan creation itself would consume significant Claude context. The Codex planner returns a structured markdown plan the caller can review, adopt, or reject.

The Boss of the product team. Veteran Head of Product who has shipped across B2B and B2C, survived pivots, and knows what wins markets. Listens, then calls it.

The Pusher of the product team. Startup PM who wants to ship the MVP yesterday, iterate on signal, and beat the competition to the punch.

The Rookie of the product team. Junior PM energy. Asks about metrics, users, and the underlying "why" behind every feature. Not contesting, just making the decision explainable.

The Watcher of the product team. Data-minded thinker who sees second-order effects, weird user behaviors, and the metric gaming everyone else missed.

Product Manager agent for product strategy, feature specifications, user stories, business logic, prioritization, and roadmap management. Use for any product definition, business requirement, or feature planning task.

Agent that analyzes user requests to create requirement specifications and WORK units. Must be used when "[]" tags are detected. For simple requirements, assumes Planner role to create PLAN.md + TASKs directly.

Strategic priority agent. Analyzes the business from a CEO perspective — growth blockers, resource allocation, build vs. buy decisions, investor-readiness. No sugar-coating. Runs in parallel with CTO/CFO/COO.

Operations execution agent. Finds what's falling through the cracks — stale work, broken processes, missing automation, communication failures. What the CEO doesn't see.

Implements code for any technology stack strictly from sprint spec files, then returns a structured IMPLEMENTATION REPORT with conformity status and deviations. Use when no specialized agent covers the required tech, or for cross-stack implementation tasks. Trigger with "implement sprint task", "build from specs".

Use when a decision needs multi-dimensional analysis — weighing intent/fit, scope/risk, and alternatives before a design or approach is chosen. The synthesis investigator behind spec analysis.

Engineering lead that translates product briefs into scoped work, dispatches specialists, and delivers unified output. Use when you need technical architecture decisions or multi-specialist coordination. Trigger with "scope this feature", "dispatch the engineering team".

Use when designing a new system or subsystem, or making a structural change to an existing one — the system-design specialist that decomposes the architecture, sets module/service/component boundaries, analyzes failure modes and data flow, designs frontend behavior at scale (debouncing, virtualization, pagination, caching), and reviews structural changes for drift. Verifies against the architect persona standards.

Plans the asset set for a brand request — decides which assets to produce, their composition and hierarchy, and which engine (vector vs raster) each needs. Use before generating anything ambiguous or multi-asset.

Write architecture docs, ADRs, API specs, onboarding guides, and system diagrams using the Diátaxis model. Use when the team needs durable, accurate technical documentation. Trigger with "write an ADR", "document this architecture".

Plans efficient batch recording sessions that group multiple videos by topic, manage energy and breaks, and produce detailed run-of-show schedules to maximize output per recording day. Use when scheduling a recording block. Trigger with \"plan batch recording session\", \"schedule my recording day\".

Design ticket workflows, SLA tiers, knowledge-base structures, and escalation paths for stage-appropriate support operations. Use when building or auditing a support operation from triage to Tier 3. Trigger with "design a support workflow", "build a support knowledge base".

Use when a task has been triaged and the chain needs to decide which specialist agents are responsible — the router consulted once after triage to finalize the reviewer/investigator roster, web-research scope, and fan-out approvals.

Generates strategic 30-day content calendars with release timing, platform mix, batch recording days, and content diversity targets to grow a creator channel consistently. Use when planning a month of content. Trigger with \"build my content calendar\", \"plan my posting schedule\".

Builds time series forecasting models for demand, revenue, and usage signals. Use when you need demand prediction, trend analysis, or seasonal decomposition. Trigger with \"forecast this time series\", \"build a demand model\".

Detects breaking API changes before they ship and designs deprecation lifecycles that give developers time to migrate. Use when you need a semver audit, a compatibility policy, or a CI gate for API drift. Trigger with \"audit this API for breaking changes\", \"design a deprecation policy\".