Install any skill in seconds. Free to start, no credit card required.
Get Started Free →lawve-ai/matter-plan-builder-scott-margetts
Convert agreed scope into a structured matter plan — phases, workstreams, milestones, dependencies, owner assignments, and matter setup decisions.
Passed 11 of 21 test cases with the skill, and 7 of 21 without it.
and 5 more in the full report
8 fixed4 passed either way7 still failing3 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
Assign workstream owners for Project Zenith: Workstream 1 (French Regulatory), Workstream 2 (German IP). Our draft lists WS1 owner as Paris Regulatory Team and WS2 owner as GNS Law Firm.
| Checked | Without skill | With skill |
|---|---|---|
| Requires named individuals rather than general team names or firm names as workstream owners. | Missed | Passed |
The instructions your model reads when the skill loads.
Convert agreed scope into a structured matter plan the team can execute against. A plan that exists only in the partner's head is not a plan — it is an intention. The function of this skill is to make the plan explicit, assign ownership, sequence the work, and produce an output that every other LPM discipline can reference.
This skill takes the output of matter-intake-scoping (or an equivalent scope description) and produces the planning layer. scope-change-controller manages that plan as the baseline throughout the matter. status-report-drafter reports progress against it. timeline-generator adds dependency logic and critical path visualisation.
The matter setup decision is part of this skill. How the matter is configured in the billing system — single matter vs phased structure, task codes, matter numbers — directly affects whether the data collected is useful for reporting, billing, and future scoping. Getting it wrong at setup costs the project throughout its lifetime. This is not a billing admin task. It is a strategic planning decision that must be made before time starts being recorded.
Scope to complete plan. Matter plan (phases, workstreams, milestones, key dependencies, owners) plus workstream plans for each workstream. Produces matter setup recommendations alongside the plan. Default mode for matters of moderate complexity with a defined team.
Mode 1 produces a workstream plan for every workstream without exception. If there are five workstreams, five workstream plans are produced. Do not produce the first workstream and note that others "follow the same format" — produce them all. Mode 2 is the correct mode when a matter plan without workstream detail is needed. If a user invokes Mode 1 on a matter with many workstreams and full workstream plans would be disproportionate, ask whether Mode 2 is more appropriate before proceeding.
Phases, workstreams, high-level milestones, and owners only. Workstream plans are the responsibility of workstream leads — this skill produces the matter plan and the template workstream leads should follow. Produces the dependency-flagged plan for timeline-generator.
A single workstream or jurisdiction plan built in detail. Input: the matter plan (or equivalent) plus the workstream scope. Output: task-level plan with sequencing, dependencies, durations, and owners. Designed for the workstream lead to own and maintain.
For matters where full scope is not yet defined. Plan the current phase in full detail. Produce a stub plan for subsequent phases — milestones only, no task detail. Flag the trigger points at which the next phase needs to be planned in detail. The stub is a placeholder, not a commitment — mark it as such.
The most frequently needed mode in practice. Accepts emails, call notes, or meeting notes and proposes updates to the existing plan — status changes, progress notes, due date revisions, new blockers, completed tasks. The LPM reviews and confirms; they do not create the updates manually.
This mode exists because the alternative — asking lawyers to update the plan directly — does not work. Lawyers do not update plans. The information exists in their emails and in their heads. The LPM's job is to extract it from those sources without creating a manual data-entry burden that takes longer than the matter itself.
Input: existing plan (uploaded as file or pasted) + correspondence since last update. Output: proposed plan changes, presented as a confirmation list. The LPM confirms, dismisses, or edits each proposed change. The updated plan is produced as a new version (.docx and structured export) on confirmation.
In connected mode, this mode can be triggered automatically — Claude monitors matter correspondence and surfaces proposed updates without waiting for the LPM to initiate.
test cases passed, vs 7 / 21 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.