Stats
Actions
Tags
From splock
Author a structured plan for a slug via the two-call planner (Call 1 free-form reasoning, Call 2 schema-valid JSON emission). Use when the user says "plan X", "write a plan for X", "design the approach to X", or after recon when an initiative needs success criteria + a task skeleton. Emits <slug>_plan.json and renders the <slug>_plan.md twin.
How this skill is triggered — by the user, by Claude, or both
Slash command
/splock:planThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Two-call planning engine. Produces a schema-valid plan substrate
Two-call planning engine. Produces a schema-valid plan substrate
(docs/plans/<slug>/<slug>_plan.json) and its rendered Markdown twin
(<slug>_plan.md).
Operator entry: /plan <slug> [free-text]. A bare slug is a fresh plan;
additional tokens carry a directive or a --reopen/--amend mode.
Routes through bin/plan -> python -m bin._planner.main plan, which makes two
distinct model calls: Call 1 is free-form reasoning, Call 2 is constrained to
schemas/plan_v1.schema.json. The planner subagent (agents/planner.md) is
the model-side contract; the driver makes the calls.
The emitted <slug>_plan.json is sealed substrate — surgical edits go through
bin/plan --amend, never a raw file edit.
npx claudepluginhub splockorg/splock --plugin splockCreates bite-sized, testable implementation plans from specs or requirements, with file structure and task decomposition. Activates before coding multi-step tasks.