Plan adoption engine

SkillDev tools

Internal Plumbline engine. Use only after Plan or Execute selects adoption of an explicit Checkpoint Relay artifact.

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use the Plan adoption engine skill

What this skill tells your AI

The instructions your AI receives, as published by nickyfactz/plumbline in skills/plumbline-plan-adoption-engine/SKILL.md and read by ahel’s review.

Outcome

Given an otherwise sufficient external specification, plan, handoff, work order, or equivalent controlling artifact, create one smallest companion live plan for deterministic Checkpoint Relay. Keep the source as authority; this is structural adoption, not another product-planning phase.

Use this engine only when the user explicitly selected checkpoint_relay and the controlling source is semantically sufficient for execution but does not satisfy the normalized Relay Execution Contract in references/checkpoint-relay.md and references/plan-schema.md. Ordinary continuous Execute remains artifact-agnostic and does not require adoption.

Classify before writing

Choose exactly one result:

  • relay_compatible: validate the existing plan without rewriting it;
  • adoptable: the source settles objective, scope, meaningful checkpoint outcomes/dependencies, and proof, but needs a companion execution record;
  • insufficient: return to normal Plan or Shape because a material contract or product decision is absent.

Do not treat missing Plumbline headings alone as semantic insufficiency. Do not normalize vague prose optimistically merely to make Relay start.

Materialize the smallest companion

For an adoptable source, keep the source unchanged and reference it by relative path plus hash. Carry only the execution/recovery structure Relay needs:

  • one objective and selected source set;
  • execution_mode: checkpoint_relay;
  • an enumerable checkpoint sequence with explicit dependencies and observable acceptance;
  • exactly one current checkpoint and matching status;
  • one legal next_safe_action that names the current checkpoint;
  • explicit approval gates, blockers, residual risks, and non-goals already in the source;
  • the compact resume and recovery fingerprint fields from the plan schema.

Preserve checkpoint meaning and useful source structure. Carry only source- supported product behavior and acceptance into one companion execution plan.

Readiness and Git boundary

Run node runtime/relay-readiness.js <plan> from the plugin root. The validator checks only normalized structure and repository durability; it never interprets arbitrary source prose. relay_ready: false blocks automatic Relay.

The main lifecycle owner reviews the companion and establishes the coherent Git recovery boundary. Stage and commit only the companion and its selected repository-local sources. Re-run readiness after that boundary; start Relay only when the plan and those sources are tracked and unchanged.

Completion

Adoption is complete when exactly one checkable result exists:

  • relay_compatible: readiness identifies the existing plan as Relay Ready;
  • adoptable: one source-referencing companion reaches relay_ready: true after the main-owned Git boundary; or
  • insufficient: no companion is created and the missing material contract or product decision is named.

Signals

GitHub stars
22
Last commit
Sep 2026
Advanced
Item type
skill
Key
plumbline-plan-adoption-engine
Source
github.com/nickyfactz/plumbline