plan

name	plan
description	Derive an Implementation Plan 2.0.0 (with Verification Delta 2.0.0) from exactly one Decision-Complete Work Brief 2.0.0.
vibe-protocol	implementation-plan-producer
vibe-input-artifact	work-brief
vibe-output-artifact	implementation-plan-with-verification-delta
runtimeExecutionClaim	pending-live

plan

Derive an Implementation Plan 2.0.0 (with embedded Verification Delta 2.0.0) from exactly one operator-locked Decision-Complete Work Brief 2.0.0. Plan derives; it decides nothing.

What this skill owns

One Implementation Plan 2.0.0 under .vibe/work/<work-id>/implementation-plan.json with the embedded Verification Delta 2.0.0. The plan carries first-class verticalSlices, dagEdges with parallel-safety rationale, pathOwnership, schematicInvocations, capabilityWork, tddSeams, migrationSequence, acceptanceTrace over every step, sourceTrace over every derived statement, noNewDecisionsProof, and an independent planValidation. The retired v1 plan extension is not produced — the v2 first-class fields supersede it.

Intake

Consume exactly one effective Work Brief 2.0.0 that is schema-valid, reviewed, operator-locked, digest-stable, and free of implementation-affecting unknowns. A structurally plausible but gapped brief opens a forward-only gap — it is never a dead-end rejection and the operator is never bounced back to brainstorm or grill-me.

Derivation

Map locked decisions into execution and verification only: vertical slices sized for focused contexts, a dependency-valid DAG, owned/read-only/untouchable paths, exact schematic invocations, capability work, migration/rollout/rollback order, acceptance-to-step-to-evidence trace, and the Verification Delta 2.0.0 with per-item source links, runner and trust class, input fingerprints, blocking and freshness and reuse policy, and machine-validated not_applicable rationales. Every derived item cites a locked decision or repository standard.

What plan may not do

Plan introduces or changes no behavior, UX, domain term, API/data, architecture, dependency, security/perf/ops objective, test seam that changes architecture, Atlas vision, scope, risk acceptance, or quality threshold. When more than one materially valid plan remains, plan emits a typed discovery_gap and routes it to the forward-only resolver — it never chooses between valid architectures and never asks the operator to restart discovery.

Approval and recovery

The operator approves the exact persisted base plan; later corrections are append-only Plan Amendments. A material choice smuggled in a Plan Amendment is reclassified as a Decision Amendment. Recovery is always forward-only inside this run. A miss is never a dead-end rejection and never instructs the operator to re-invoke an earlier skill. The current run launches the forward-only gap resolver, persists a scoped Decision or Plan Amendment bound to the exact amendment digest, revalidates affected closure with targeted invalidation of only the impacted evidence, and resumes automatically. When bounded in-run resolution cannot establish a safe path, the run records an honest resumable wait or hard block with an exact next action — it never converts that into prose-green and never bounces the operator back to an earlier skill as recovery.

Artifact carrier

• Persist this skill's one output under .vibe/work/<work-id>/ as a UTF-8 JSON or Markdown artifact; chat history alone is never a carrier.
• Verification evidence lives under .vibe/evidence/<work-id>/; the starter runner catalog is .vibe/registry/runner-catalog.json.
• Runtime execution claim: pending-live — native pi loading and execution stay unproven until a recorded authenticated runtime witness exists.

Forbidden outputs

This skill never produces: build-result, ship-packet, push, pull-request. Recovery stays forward-only inside this run; a miss is resolved in place through the current-run gap resolver, never by handing the operator back to an earlier skill.