EJ

ejs-session-wrapup

Automates the creation and maintenance of session-long tracking artifacts and documentation.

Install

mkdir -p .claude/skills/ejs-session-wrapup && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/15298" && unzip -o skill.zip -d .claude/skills/ejs-session-wrapup && rm skill.zip

Installs to .claude/skills/ejs-session-wrapup

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Initialize and manage EJS artifacts throughout session lifecycle (Session Journey initialization, continuous updates, and finalization; ADR only when significant)
162 chars✓ has a “when” trigger
Intermediate

Key capabilities

  • Initialize session journey artifacts
  • Continuously update session journey during work
  • Finalize session journey at session end
  • Generate unique session IDs
  • Draft Architecture Decision Records (ADRs) when criteria are met

How it works

The skill manages the session lifecycle by creating an initial Session Journey artifact, updating it with interactions and learnings, and finalizing it with an optional ADR based on decision criteria.

Inputs & outputs

You give it
User signals (start/end session, task details), agent interactions
You get back
Session Journey artifact, optional Architecture Decision Record (ADR)

When to use ejs-session-wrapup

  • Tracking progress on GitHub issues
  • Starting a new feature session
  • Finalizing work with an ADR

About this skill

Engineering Journey System — Session Journey Management Skill

Purpose

Automate Session Journey management throughout the entire session lifecycle.

This skill handles three phases:

  1. Session Initialization - create initial Session Journey artifact at session start
  2. Continuous Updates - update Session Journey throughout the session as work progresses
  3. Session Finalization - finalize Session Journey and optionally create ADR at session end

This skill integrates with GitHub Copilot or compatible AI coding agents.

Triggers

Session Start Triggers

Activate session initialization when the user signals session start with any of these phrases:

  • "Let's start"
  • "Begin"
  • "Start working on [task/issue]"
  • "Initialize session"
  • "Create session journey"

Also treat these as session-start signals when beginning new work:

  • Starting work on a GitHub issue
  • Beginning a new feature or bug fix
  • "Let's work on..."

Session End Triggers

Activate session finalization when the user signals session end with any of these phrases:

  • "Generate ADR for this session"
  • "Wrap up this session"
  • "End session"
  • "Finalize journey"

Also treat these as session-end signals:

  • "I'm done"
  • "Ship it"
  • "Commit this"
  • "Commit and push"
  • "Push this"

Actions on Session Start

1. Switch to Session Initialization Mode

  • Prepare to create the initial Session Journey artifact
  • Establish session context

2. Generate Session ID

  • Format: ejs-session-YYYY-MM-DD-<seq>
  • Ensure uniqueness per day

3. Create Initial Session Journey

  • Use ejs-docs/journey/_templates/journey-template.md as base
  • Populate initial metadata:
  • session_id
  • author
  • date
  • repo
  • branch
  • agents_involved
  • decision_detected: false (initial state)
  • Fill in initial sections:
  • Problem/Intent - capture the goal or task being addressed
  • Interaction Summary - create structure for continuous updates
  • Leave other sections ready for incremental updates during the session
  • Write to: ejs-docs/journey/YYYY/<session-id>.md

4. Inform User

  • Confirm Session Journey created
  • Remind that journey will be updated throughout session
  • State session ID for reference

Actions During Session (Continuous Updates)

As the session progresses, continuously update the Session Journey with:

Update Interaction Summary

  • Add new human prompts and agent responses
  • Record outcomes and decisions made
  • Keep format: - Human: <prompt> → Agent: <response> → Outcome: <result>

Update Experiments Section

  • Record what was tried
  • Document what happened
  • Note what evidence emerged

Update Iteration Log

  • Document pivots and refinements
  • Explain why changes were made

Update Decisions Made

  • Capture decisions with rationale and impact
  • Assess if decision meets ADR criteria

Update Key Learnings

  • Record insights as they emerge (technical, prompting, tooling)

Trigger ADR Creation If Needed

  • If a significant decision occurs mid-session, note it in the journey
  • Prepare ADR metadata for later creation if decision rubric is met

Actions on Session End (Finalization)

1. Switch to Journey Finalization Mode

  • Stop suggesting new code or expanding scope
  • Focus on completing the Session Journey

2. Finalize Session Journey

  • Review and complete all sections:
  • Ensure Interaction Summary is complete and coherent
  • Complete Agent Collaboration Summary
  • Complete Agent Influence section
  • Finalize Experiments/Evidence
  • Complete Iteration Log
  • Finalize Decisions Made
  • Complete Key Learnings
  • Fill in "If Repeating This Work"
  • Complete Future Agent Guidance
  • Populate all ## MACHINE EXTRACTS sections:
  • INTERACTION_EXTRACT
  • DECISIONS_EXTRACT
  • LEARNING_EXTRACT
  • AGENT_GUIDANCE_EXTRACT
  • Update decision_detected field based on ADR criteria
  • Save final version to ejs-docs/journey/YYYY/<session-id>.md

3. Decision Detection (ADR Gate)

Create an ADR only if at least one applies:

  • System boundary changed (service/datastore/topology/major dependency)
  • Public contract changed (API/schema/config/capability)
  • Security/privacy/compliance posture changed
  • Credible alternatives weighed with meaningful trade-offs
  • Long-lived or hard-to-reverse consequences
  • Process/workflow change affecting future work

If none apply: set decision_detected: false in the Session Journey and do not create an ADR.

4. Draft ADR (Only If Needed)

  • Use ejs-docs/adr/0000-adr-template.md as base
  • Create a numbered ADR at: ejs-docs/adr/NNNN-<kebab-title>.md
  • Requirements:
  • include session_id
  • include a link back to the Session Journey file (relative link)
  • record considered options + rationale + consequences
  • keep it decision-focused; details live in the Session Journey
  • Update Session Journey adr_links field with new ADR path

Notes for Agents

Throughout Session

  • Keep Session Journey up-to-date as work progresses
  • Don't wait until the end to capture important context
  • Update incrementally to preserve accurate history
  • Treat the journey as a living document during the session

At Session End

  • Treat session content as frozen once end-session is triggered
  • Ensure all decisions and learning points are explicit
  • Do not overwrite previous Session Journeys or ADRs with the same session ID
  • Links between ADR and Session Journey should be relative

Benefits of Incremental Capture

Capturing the journey throughout the session (vs. reconstructing at the end):

  • Better context preservation - details captured when fresh
  • More accurate collaboration trail - prompts and responses recorded as they occur
  • Reduced end-of-session burden - most work already done
  • Higher quality documentation - real-time vs. retrospective capture
  • Better for multi-step/multi-agent sessions - preserves full collaboration history

Reminder: skills don't auto-run on git events

Skills are loaded by prompt relevance (based on this file's description and your request). For "fire on commit/push" enforcement/reminders, use the git hooks in this repo.

When not to use it

  • When the user wants to overwrite previous Session Journeys or ADRs with the same session ID
  • When the user wants to reconstruct the journey at the end of the session
  • When the user wants to generate an ADR for every decision

Limitations

  • ADR creation is conditional on specific criteria.
  • The skill does not overwrite previous Session Journeys or ADRs with the same session ID.
  • Session content is frozen once an end-session trigger is activated.

How it compares

This skill captures the session journey incrementally and in real-time, providing a more accurate and detailed historical record than reconstructing it retrospectively.

Compared to similar skills

ejs-session-wrapup side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
ejs-session-wrapup (this skill)06moNo flagsIntermediate
task-management03moNo flagsBeginner
pi-tasks03moNo flagsIntermediate
self-host-development-light02moNo flagsBeginner

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

task-management

rjchien728

管理 ~/tasks/(或使用者指定路徑)下的進行中工作項目資料夾。每個 task 是一個資料夾,README.md 是純狀態 dashboard(status frontmatter + 一句 headline + 一行狀態 + 檔案清單 + 下一步),實際的設計 / 規劃 / 調查文件用獨立 .md 檔,腳本與 reproduce 也存獨立檔。負責命名(YYMMDD-<slug>)、建資料夾、寫 README dashboard、更新狀態、總覽掃描。當使用者說「開新 task」「開個 task」「新增 task」「new task」「記錄到 tasks 下」「把目前設計記下來開個 task

00

pi-tasks

Micka33

>-

00

self-host-development-light

doolin

Lightweight self-hosted project management using markdown files in a .development directory. Covers backlog, planning, todo, roadmap, and saved plans — everything an agent needs to orient and start working.

00

working-summary

Innei

Use when the user asks for a work summary, weekly report, 工作总结,周报,working summary, or sprint/period recap. Aggregates GitHub PR/commit/issue activity from configured repos, optionally reads Linear cycle issues via MCP, honors Chinese public holidays, and produces a markdown report. Default range is

00

add-todo

MichaelGao1999

即时写入待办项到 status.md,当用户说"计入待办""记入待办""加入待办""加到清单""后面做""回头处理"时使用。

00

yatta

Trompetilla

Personal productivity system for task and capacity management. Create and organize tasks with rich attributes (priority, effort, complexity, tags), track time and streaks, manage capacity across projects and contexts, view Eisenhower Matrix prioritization, sync calendar subscriptions, handle delegat

00

Search skills

Search the agent skills registry