customize-from-assessment

Customize from Assessment

This is the first-run skill for NightShiftOS. It takes the user's assessment output and transforms a generic plugin install into a personalized business operating system: rewritten root CLAUDE.md, a generated personal voice-rules skill, and a tailored set of active skills.

Run this exactly once per install. Re-run only when the user retakes the assessment or asks to refresh from a new output.

When this skill triggers

The user has already completed the assessment at nightshift.jaredlangley.cc and has an assessment-output.md block to paste. Common phrasing:

• /customize-from-assessment
• "I just took the assessment"
• "customize this plugin for me"
• "set up my business OS"
• "here's my assessment output"

If the user invokes this skill without an assessment output, route them to https://nightshift.jaredlangley.cc and stop. Do NOT fabricate assessment data.

The customization flow

Execute in order. Do not skip steps.

Step 1: Accept the assessment block

Prompt the user: "Paste your assessment output below. It starts with NIGHTSHIFTOS ASSESSMENT OUTPUT and ends with END ASSESSMENT."

Validate the block has:

• Header line NIGHTSHIFTOS ASSESSMENT OUTPUT
• 12 sections labeled ## 01 Identity through ## 12 Coach Style
• Trailing line END ASSESSMENT

If malformed, ask for a re-paste. Do not proceed with partial data.

Step 2: Parse into structured profile

Extract these fields from the 12 sections:

identity: { name, roles[], one_line }
values: { picked[], written[], protected_time[] }
brand_state: { status, has_assets, target_feel }
voice: { samples[], banned_phrases[], signature_phrases[], tone_ratings{}, archetype[] }
current_state: { stage, hours_per_week, active_clients, runway_months }
offer: { what, who, price, unit_econ }
sales_motion: { inbound_pct, outbound_pct, channels[], cycle_days, avg_deal }
marketing: { active_channels[], want_channels[], cadence, working[] }
goals: { ninety_day, twelve_month, why }
risk: { free_first_level, max_tool_spend, max_client_hours, worst_case }
stack: { tools[{function, name, cost}], total_monthly }
coach_style: { challenge_vs_affirmation, nudge_cadence, hard_vs_gentle }