Convert agreed scope into a structured matter plan — phases, workstreams, milestones, dependencies, owner assignments, and matter setup decisions. Use when planning a new matter, running a kickoff, building a workstream plan, structuring phases, setting up task codes, or producing a plan to drive status reporting. Trigger on: 'build a plan', 'matter plan', 'project plan', 'what are the phases', 'workstream plan', 'how do we sequence this', 'who owns what', 'task codes', 'matter setup', 'workstream plan', 'matter plan', 'rolling wave', 'plan the next phase', 'what comes first', 'kickoff agenda'.
56
63%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/matter-plan-builder/SKILL.mdQuality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong, well-crafted skill description that clearly defines its domain (legal matter planning), lists specific concrete deliverables, and provides extensive trigger guidance. The explicit 'Use when' clause and 'Trigger on' list ensure Claude can reliably select this skill. The only minor issue is a duplicate entry ('workstream plan' and 'matter plan' each appear twice in the trigger list), but this doesn't materially affect quality.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: converting scope into a structured matter plan, phases, workstreams, milestones, dependencies, owner assignments, and matter setup decisions. These are concrete, domain-specific deliverables. | 3 / 3 |
Completeness | Clearly answers both 'what' (convert scope into structured matter plan with phases, workstreams, milestones, dependencies, owner assignments, matter setup decisions) and 'when' (explicit 'Use when...' clause plus a detailed 'Trigger on:' list with specific phrases). | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms including 'build a plan', 'project plan', 'what are the phases', 'who owns what', 'kickoff agenda', 'how do we sequence this', and domain-specific terms like 'task codes', 'rolling wave', 'matter setup'. These reflect how users would naturally phrase requests. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche in legal matter planning. Terms like 'matter plan', 'task codes', 'matter setup', and 'workstream plan' are domain-specific enough to avoid conflicts with generic project management or document skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
27%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill demonstrates deep domain expertise in legal project management and provides genuinely useful structural definitions (field schemas, dependency types, operating modes). However, it is severely over-length, spending significant token budget on editorial justification and concept explanation rather than concise instruction. The lack of concrete output examples and the monolithic structure significantly reduce its effectiveness as a skill file.
Suggestions
Reduce content by 50-60%: remove editorial commentary (e.g., 'A plan that exists only in the partner's head is not a plan'), concept explanations Claude already knows (what milestones are, what dependencies are), and the extended justification for why Mode 5 exists. Keep only the instruction.
Split into multiple files: move domain knowledge (phase patterns) to PHASE_PATTERNS.md, standard plan fields to PLAN_SCHEMA.md, common planning failures to PLANNING_ANTIPATTERNS.md, and reference them from the main skill with one-line descriptions.
Add a concrete worked example: show a small but complete example plan output (even for a 2-workstream matter) demonstrating the exact format, field population, and structured export so Claude can pattern-match rather than interpret prose descriptions.
Add a post-production validation checklist at the end of Step 6: verify all workstreams have plans, all required columns are present, all predecessor references resolve to valid Task IDs, all milestones have owners and dates, and structured export matches the .docx content.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~400+ lines. Extensive sections explain concepts Claude already understands (what phases are, what milestones are, why lawyers don't update plans, what dependencies are). The 'Common Planning Failures' section is largely editorial commentary rather than actionable instruction. The explanation of why Mode 5 exists spans multiple paragraphs of justification rather than simply instructing. Many paragraphs could be reduced to single sentences. | 1 / 3 |
Actionability | Provides detailed field schemas, table structures, and clear output format specifications which are concrete and useful. However, there is no executable code, no actual template content, no example plan output showing what a completed matter plan looks like, and no example of a Mode 5 update confirmation list. The guidance is specific in structure but lacks concrete worked examples that would make outputs copy-paste ready. | 2 / 3 |
Workflow Clarity | The 6-step process is clearly sequenced with good detail on what each step produces. Step 1 includes a validation checkpoint (confirming owner names before proceeding). However, there are no explicit validation steps after plan production (e.g., verify all workstreams have plans, verify all required fields are populated, verify cross-references are consistent). For a skill producing complex structured outputs, the absence of a final validation checklist is a gap. | 2 / 3 |
Progressive Disclosure | Monolithic wall of text with no references to external files despite the content clearly warranting separation. The domain knowledge sections (phase patterns, common failures), standard plan fields, and communication rhythm sections could each be separate reference files. The entire document is ~400+ lines of inline content with no bundle files to support it, making it overwhelming to consume in a single read. | 1 / 3 |
Total | 6 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
dc6509d
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.