skills.rest
yang-l
yang-lCOMMUNITY
Explore all skills by this author →
GitHub

wrapup

Ends work sessions with two grounding questions that surface concrete risks and blind spots.

Updated Jul 13, 2016
Search Tags:#retrospective#handoff#blind spots#risk assessment#self-audit#session review#wrap up
One-click install
npx skills add https://github.com/yang-l/configs --skill wrapup-yang-l
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: wrapup
Source: https://github.com/yang-l/configs/tree/main/.claude/skills/wrapup
Command: npx skills add https://github.com/yang-l/configs --skill wrapup-yang-l

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Sessions often end with unverified assumptions, silent risks, and missed blind spots that only surface later as bugs or rework. This Skill runs a deliberate end-of-session ritual that forces specific, evidence-based answers to two grounding questions instead of generic reassurances. ## Core Features & Use Cases - Confidence audit: Lists 4-8 concrete uncertainties tied to actual files, decisions, and assumptions from the session, each with what could go wrong. - Risk flagging and follow-through: Marks items as HIGH RISK when they involve silent wrong behavior, unverified architectural assumptions, security gaps, or irreversible actions, and offers to investigate them before closing. - Outside-view check: Applies Sam Altman's question through five lenses (framing, unstated assumptions, scope drift, simpler alternatives, unraised risks) to catch what neither party has said out loud. - Use Case: After a long refactoring session, trigger the wrap-up to discover that a renamed function was only grepped in one directory and that the session optimized the wrong part of a slow page. ## Quick Start Say "let's wrap up this session" to run the two-question end-of-session review before closing out your work.

Frequently Asked Questions about wrapup

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run an end-of-session review with an AI assistant?▼

Trigger the wrap-up ritual by saying "let's wrap up", "before we end this session", or "end of session check". The assistant then answers two grounding questions built from the session's actual files, decisions, and assumptions rather than generic advice.

What questions should I ask before ending a work session?▼

Ask "What are you least confident about right now?" for an introspective audit of the session's work, and Sam Altman's "What's the biggest thing I'm missing about the situation right now?" for an outside view. Each targets a different vantage point so answers do not overlap.

How does the wrap-up skill decide which risks to flag?▼

Items are marked HIGH RISK only if they could cause silent wrong behavior, rest on unverified architectural assumptions, involve security-relevant gaps, or involve unverified destructive actions. Minor or cosmetic items stay unmarked so the flag stays meaningful.

When should I not trigger the session wrap-up?▼

Do not trigger it on standalone mid-task confidence remarks like "I'm not confident about X". It is a deliberate end-of-session ritual, not a running commentary tool, and only fires when you are actually closing out work.

What happens if the wrap-up finds a high-risk item?▼

The assistant states plainly that the item warrants checking before ending and explicitly offers to investigate it immediately, such as reading the relevant code, tracing the actual behavior, or asking the one clarifying question that resolves it.

Related Skills

Explore More Software Engineering Skills→
jstoup111jstoup111
5

architecture-review

Review PRDs and technical intent for architectural alignment before coding.

Community
Advanced
Compass-BrandCompass-Brand

strategic-compact

Generate structured session handoff documents before context limits are reached.

Official
Basic
tobihagemanntobihagemann
399

review-code

Run parallel code, security, and peer reviews with automated fixes and test verification.

Community
Advanced
RefoundAIRefoundAI
1.3k

running-meetings

Guides facilitation of decision-oriented meetings using frameworks from Lenny's Podcast and Newsletter.

Official
Intermediate