cold-start-interview

Cold-start setup — learns your jurisdictional footprint and escalation rules from your handbook and termination memos. Asks which states and countries have employees, reads seed documents, and builds a jurisdiction-aware escalation table. Use on fresh install, when CLAUDE.md still has [PLACEHOLDER] markers, or when re-running with --redo or --check-integrations.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

77%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A highly actionable, well-sequenced interview skill with strong validation checkpoints and concrete formats. Its weaknesses are verbosity (repeated reassurance and orientation prose) and a monolithic structure that offloads nothing into reference files despite its length.

Suggestions

Trim repeated reassurance — consolidate the 'you can change anything later' and 'Why this matters' passages into single statements; the close-out change-later note appears in near-duplicate form across the After-writing and Close sections.

Move the long interview scripts, orientation paragraphs, and example phrasings into a references/ file (e.g. INTERVIEW-SCRIPT.md) and link one level deep, so SKILL.md stays a lean overview with the step sequence and validation rules.

In the description, add natural-language triggers ('set up', 'configure', 'onboard the employment-legal plugin') alongside the existing flag-based ones so users who don't know the --redo/--check-integrations flags can still surface the skill.

Dimension	Reasoning	Score
Conciseness	At ~300 lines the body is mostly purposeful interview script and branch logic, but it is padded with repeated reassurance ('change anything later' restated in several places), a verbose 'Why this matters' pitch, and lengthy orientation prose that could be tightened; not a 1 because it avoids explaining concepts Claude already knows.	2 / 3
Actionability	Provides concrete file paths, exact marker formats ('<!-- SETUP PAUSED AT: -->', '[PENDING]' vs '[PLACEHOLDER]'), example jurisdiction-table rows (California, New York), and a precise ✓/⚪/✗ integration-reporting format with branch logic; fully actionable guidance rather than vague direction.	3 / 3
Workflow Clarity	Sequences a clear multi-step flow (6-step summary, Parts 0–3, build table, write profile, close) with explicit validation checkpoints — verify user-stated legal facts, review skipped questions before writing, test MCP connections before reporting ✓, and pause/resume mechanics; feedback loops are present so it is not capped at 2.	3 / 3
Progressive Disclosure	Well-sectioned but monolithic: no references/scripts/assets bundle files exist, and large blocks that could live in reference files (full interview scripts, orientation phrasing, example close-out notes) are inline; the only referenced paths point to the plugin root, not skill bundles. Not a 3 because there are no one-level-deep skill-bundle references; not a 1 because sectioning is clear rather than a disorganized wall.	2 / 3
	Total	10 / 12 Passed

Description

85%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, concrete description that clearly states both capability and explicit use-conditions, scoped tightly to employment-legal onboarding. Its only weakness is trigger phrasing that relies on command flags and jargon ('cold-start') rather than the natural language a user would actually say.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'learns your jurisdictional footprint and escalation rules', 'Asks which states and countries have employees', 'reads seed documents', and 'builds a jurisdiction-aware escalation table' — matching the multiple-specific-actions anchor; not a 2 because it goes beyond naming a domain plus a few actions.	3 / 3
Completeness	Clearly answers both what (learns footprint/escalation rules, builds escalation table) and when ('Use on fresh install, when CLAUDE.md still has [PLACEHOLDER] markers, or when re-running with --redo or --check-integrations'), matching the explicit-trigger anchor; not a 2 because the when-clause is explicit, not implied.	3 / 3
Trigger Term Quality	Contains relevant domain terms (handbook, termination memos, escalation rules) but the when-triggers lean on command flags ('fresh install', '--redo', '--check-integrations', 'cold-start') rather than natural user utterances like 'set up' or 'configure'; missing common variations keeps it at 2 rather than 3.	2 / 3
Distinctiveness Conflict Risk	Scoped specifically to employment-legal cold-start setup with jurisdictional-footprint triggers unlikely to fire for other skills; clear niche, low conflict risk.	3 / 3
	Total	11 / 12 Passed

Validation

81%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 2 missing	Warning
referenced_paths_exist	Referenced path issues: 1 missing	Warning

	Total	13 / 16 Passed

Repository: anthropics/claude-for-legal
Path: employment-legal/skills/cold-start-interview/SKILL.md
Commit: 4a6c651

Reviewed: 6 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

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.