KBD Process
24 skills. Source of truth: skills/.
kbd-analyze
Use to run the Analyze stage of the KBD lifecycle (between Assess and Spec) — research the engineering landscape (existing open-source libraries, frameworks, skeletons that fit the assessed gaps, plus stack discovery when none is specified) and write the candidate set that the Spec and Plan stages consume.
kbd-apply
KBD-owned spec-apply driver. Wraps a spec backend (openspec; native-kbd as the always-available fallback) and drives it ONE task at a time, so KBD stays the source of truth — every task boundary fires KBD hooks, emits a plain-text position signal, and syncs progress.json and the waypoint. Replaces the broken pattern of handing the turn to a bare "implement everything" command that runs outside KBD.
kbd-assess
Use when starting or resuming the KBD lifecycle for a project — inspects the current codebase against the active phase's goals and produces a structured gap report at .kbd-orchestrator/phases/<phase>/assessment.md. Project-agnostic: reads AGENTS.md, spec files, and the codebase itself. Also reads progress.json to account for cross-tool work done since the last session.
kbd-audit
Use for a read-only causal audit of the active KBD run — exact position, lifecycle history, plan revision, ownership, blockers, and uncommitted work — without mutating anything.
kbd-bottleneck-detector
Evaluate or repair canonical KBD task, phase, and ZeeSpec boundaries. Use when progress receipts, projections, or build gates may be stale, or when the user mentions "bottleneck detector". Do NOT use for creating or advancing phases (see kbd-new-child and kbd-next-phase).
kbd-cancel
Use to gracefully cancel the active KBD run while preserving its checkpoint and immutable audit history. Requires a non-empty cancellation reason. Cancellation is terminal.
kbd-child-exit
Exit the active KBD child loop: write its handoff-out, roll its progress up to the parent node, pop the position path, and return control to the parent. The --enter companion descends into a selected child so new children nest under it.
kbd-evolve
Use for domain-landscape-first evolution of a KBD project — when the roadmap is empty, exhausted, or you want to recalibrate against external reality rather than follow internal plans. Surveys the external landscape of the project's problem domain, scores improvement opportunities against configurable criteria, and produces a ranked evolution brief that a new phase can consume as its seed.
kbd-execute
Use to select an execution backend for the active KBD phase, write canonical phase execution state, dispatch the phase, and maintain KBD as the source of truth. Supports multi-tool handoff via the progress.json protocol.
kbd-goal-check
Use after each execution turn within a goal-driven KBD phase to evaluate whether the active phase's stopping condition has been met, without the implementer agent grading its own work — returns PASS (with evidence) or CONTINUE (with the next action). Implements the maker-is-not-evaluator pattern for harnesses whose loop is a queue rather than a condition-based loop.
kbd-init
Use once per project, before any other KBD command — auto-discovers project identity, stack, and constraints from existing context files (AGENTS.md, CLAUDE.md, README.md, package.json, Cargo.toml, pyproject.toml, etc.) and generates .kbd-orchestrator/project.json and .kbd-orchestrator/constraints.md.
kbd-inject-agent-rules
Idempotently inject the agent-rules or UI/UX-routing managed pack into a target project's CLAUDE.md and/or AGENTS.md. Re-runnable — overwrites only the selected fenced region; everything else is byte-preserved. Supports --refresh to re-validate cached source URLs, and --dry-run for a target-file diff preview.
kbd-memory-recall
Query surreal-memory for prior similar KBD work and write a markdown digest at .kbd-orchestrator/phases/<phase>/prior-context.md. Used as planning input before /kbd-assess. Degrades gracefully when the memory endpoint is unreachable.
kbd-new-child
Create a child phase inside the currently-active node (arbitrary depth). Mirrors /kbd-new-phase but writes into phases/<parent>/children/<child>/, appends the new child to childPhases[], moves childPointer to it, and fires child:before. Use to split a parent phase into scoped sub-processes.
kbd-new-phase
Manually create a new top-level KBD phase. Accepts <name> [goals…] and initialises the phase directory, waypoint, project.json activePhase, and fires phase:before. Use this when no prior reflection exists (e.g. very first phase of a new project), when pivoting away from /kbd-next-phase's suggestion, or when initialising state by hand.
kbd-next-child
Advance childPointer to the next entry in childPhases[], or jump to a named child directly. Fires child:after for the closing child and child:before for the new active child. Refuses to advance past the last child (suggesting /kbd-reflect + /kbd-next-phase instead).
kbd-next-phase
Continue to the next KBD phase, automatically seeded from the previous phase's reflection. Reads the "Recommended Next Phase" section of reflection.md, initializes the new phase directory with goals.md and a skeleton progress.json, updates current-waypoint.json to point to /kbd-assess, and updates project.json activePhase.
kbd-pause
Use to gracefully pause the active KBD run, checkpoint its exact position, and prevent every harness from steering execution until an operator resumes it. Operator intent always outranks agent continuation.
kbd-plan
Use to create a prioritized, ordered change list for the current KBD project phase. Project-agnostic — reads the assessment and project constraints to produce an ordered change list, auto-detecting OpenSpec availability and emitting the appropriate change format.
kbd-process-orchestrator
Use to coordinate the full KBD (Knowledge-Based Development) lifecycle for any project — Assess, Analyze, Plan, Execute, Reflect — at every granularity level (global phases, spec-backed changes, and artifact-level QA). This is the coordination and reference document for the kbd-* sub-skills; it documents the progress.json schema, the waypoint contract, the hook taxonomy, and the sub-skill list. It has no executable logic of its own — read it once, then invoke the specific kbd-<stage> skill for the work itself.
kbd-reflect
Use to generate the phase reflection report after all changes in a KBD phase are complete — goal achievement, delivered changes, artifact quality summary, technical debt, lessons captured, and recommended focus for the next phase. Seeds the next phase's planning inputs.
kbd-resume
Use to resume a paused KBD run after validating its checkpoint and plan revision. A normal assistant response is never a resume signal — resuming requires this explicit skill.
kbd-spec
Use to run the Spec stage of the KBD lifecycle (between Analyze and Plan) — turn an assessment and analysis into concrete, ordered OpenSpec changes for KBD task execution.
kbd-status
Use to show current KBD process status — active phase, change inventory, goal completion, and next recommended action. Reads progress.json to surface work completed by all tools (Antigravity, Roo, Cursor, Cline, Codex, etc.).