Ccgs Save Session logo

Ccgs Save Session

Community
MinhLN1234
ccgs-save-session

Save the current work session to production/session-state/active.md so the next Cowork session can resume with full context: active story, what worked with evidence, what failed and why, decisions, blockers, exact next step. Use at the end of a work session, before a long session gets too heavy, or when the user says 'save session', 'lưu session', 'chốt phiên', 'wrap up for today'.

Overview

PublisherMinhLN1234
RepositoryClaude-Cowork-Game-Studio
Skill nameccgs-save-session
Stars
79
Forks
31
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by MinhLN1234 on GitHub. Read the source before you install it.

Installation

Install the Ccgs Save Session AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/MinhLN1234/Claude-Cowork-Game-Studio/tree/master/ccgs-save-session
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/MinhLN1234/Claude-Cowork-Game-Studio.git /tmp/Claude-Cowork-Game-Studio
mkdir -p .claude/skills
cp -r /tmp/Claude-Cowork-Game-Studio/ccgs-save-session .claude/skills/ccgs-save-session
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ccgs Save Session in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Ccgs Save Session on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Ccgs Save Session is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Cowork Notes (read first)

Adapted from ECC's /save-session command (github.com/affaan-m/ECC, MIT) for the CCGS pipeline on Claude Cowork. Written natively for Cowork, so there is nothing to substitute:

  1. Storage: everything lives in the user's connected project folder, never in ~/.claude/. The live file is production/session-state/active.md. Superseded snapshots go to production/session-state/archive/.
  2. Shared file: other CCGS skills append blocks to active.md (for example ccgs-story-done appends ## Session Extract — /story-done [date], the game-designer role appends section progress). This skill must carry those blocks into the snapshot, never silently drop them.
  3. Pairs with: ccgs-resume-session reads what this skill writes. The Cowork Adaptation Notes in every other CCGS skill ask for active.md to be updated at the start and end of each session; this skill and its pair are how that happens.
  4. Hooks: none. This skill is invoked by hand.

Save Session

Capture what happened in this session (what was built, what worked, what failed, what is left) so the next session starts from the real state instead of from memory. The file is written for a future session with zero memory of this one.

Output: an updated production/session-state/active.md plus an archived copy of the previous version.

Phase 1: Gather Context

Collect, from the conversation and from the project folder, before drafting anything:

  1. Active story and sprint. Read the existing production/session-state/active.md if it exists. Read the most recent file in production/sprints/ and production/sprint-status.yaml if present. Identify which story or GDD section this session worked on.
  2. Files touched. List every file created or edited this session. Read each one's current state rather than recalling it; a file that was "almost done" an hour ago may have been changed since.
  3. Evidence. For every claim that something works, find the evidence: a test that passed (name it), a value checked against a GDD section (cite it), a scene the user confirmed in Play mode. No evidence means it goes under "Not tried yet" or "In progress", not under "Worked".
  4. Failures. Every approach that was tried and abandoned, with the exact reason (error message, wrong value, design conflict). This is the most important section: without it the next session will retry the same dead end.
  5. Decisions. Design or architecture choices made in conversation that are not yet written into a GDD or ADR. Flag each one with where it should eventually live.
  6. Appended blocks. Collect every ## Session Extract or other block appended to active.md by other skills since the last snapshot.

If the session had no real work (only questions answered), say so and ask whether the user still wants a snapshot. Do not write an empty file.

Phase 2: Draft the Snapshot

Draft the file in this exact format. Write every section. If a section genuinely has no content, write the stated fallback line; an honest empty section is better than a missing one.

markdown
# Session State

**Last saved:** [YYYY-MM-DD HH:MM, user's local time]
**Project:** [project name from design/ or the folder name]
**Topic:** [one line: what this session was about]
**Active story:** [story file path and title, or "None (design work)" / "None"]
**Sprint:** [sprint file path, or "None"]
**Review mode:** [contents of production/review-mode.txt, or "lean (default)"]

---

## What We Are Building

[1 to 3 short paragraphs. The feature, system, or GDD section; why it matters; how it fits the systems index. Enough for someone with no memory of this session.]

## What Worked (with evidence)

- **[item]**: confirmed by [test name / GDD section and value / user confirmation in Play mode]

Fallback: "Nothing confirmed working yet."

## What Did NOT Work (and why)

- **[approach]**: failed because [exact reason or error]

Fallback: "No failed approaches this session."

## Not Tried Yet

- [concrete approach or idea, specific enough to act on]

Fallback: "No untried approaches identified."

## Current State of Files

| File | Status | Notes |
| ---- | ---- | ---- |
| `path/to/file` | Complete / In progress / Broken / Not started | [what is done, what is left] |

Fallback: "No files modified this session."

## Decisions Made (not yet in a GDD or ADR)

- **[decision]**: reason [why] → should be recorded in [GDD section / ADR / nowhere, conversational only]

Fallback: "No new decisions."

## Blockers and Open Questions

- [blocker or unanswered question, with who or what can unblock it]

Fallback: "No active blockers."

## Exact Next Step

[One concrete action, precise enough that resuming needs zero thinking about where to start. Include the file path and the skill to run if one applies, for example "Run ccgs-story-readiness on production/epics/combat/story-003-dash.md".]

Fallback: "Next step not determined. Review Not Tried Yet and Blockers before starting."

## Carried Forward Extracts

[Every block other skills appended since the last snapshot, copied verbatim, oldest first.]

Fallback: "None."

Phase 3: Show, Confirm, Write

  1. Show the full draft in the conversation.
  2. Ask: "May I save this to production/session-state/active.md? The current version will be archived first." Wait for a yes. Apply any corrections the user gives and show the changed sections again.
  3. On yes:
    • If active.md exists, copy it unchanged to production/session-state/archive/[YYYY-MM-DD-HHMM].md (create the folder if needed). Never append to an archived file and never overwrite one; if the name already exists, add -2, -3.
    • Write the new snapshot to production/session-state/active.md.
  4. Re-read active.md and confirm every section header is present. If anything is missing, fix it before reporting.
  5. Report: "Session saved. Previous state archived to [path]. Next session: run ccgs-resume-session."

If the user declines, write nothing. Verdict: BLOCKED, user declined write.

Rules

  • Evidence or it did not work. Do not move an item into "What Worked" because it looks right.
  • Do not record a decision as made if the user only discussed it. "Leaning towards X" goes under Open Questions.
  • Do not paste whole files or long logs into the snapshot. Point to the file path and summarize.
  • Numeric values (damage, cooldowns, wave counts) written into the snapshot must match the GDD or be flagged as "not yet in GDD". The snapshot must not become a second source of truth for balance values.
  • If the user asks to save mid-session, save what is known and mark in-progress items clearly. Saving more than once per session is fine; each save archives the previous snapshot.

Frequently asked questions

What does the Ccgs Save Session AI skill do?

Save the current work session to production/session-state/active.md so the next Cowork session can resume with full context: active story, what worked with evidence, what failed and why, decisions, blockers, exact next step. Use at the end of a work session, before a long session gets too heavy, or when the user says 'save session', 'lưu session', 'chốt phiên', 'wrap up for today'.

Why use Ccgs Save Session on TypingMind?

Because you install it once and use it with any model. Ccgs Save Session is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Ccgs Save Session in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/MinhLN1234/Claude-Cowork-Game-Studio/tree/master/ccgs-save-session. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Ccgs Save Session?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Ccgs Save Session?

As many as you like. As long as a model supports skills, you can use Ccgs Save Session with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Ccgs Save Session AI skill free?

It is published on GitHub by MinhLN1234. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇