Context Graph Compressor

Compress any conversation into a portable JSON context graph for handoff to a new chat or LLM. Trigger on: "compress this chat", "save context", "new chat", "hand off", "transfer context", "too many tokens", "start fresh", "export conversation", "context graph", "continue in new session". Also trigger proactively when the user mentions starting over in a long session.

35 stars 9 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add Adityapal67/context-graph-compressor --skill context-graph-compressor

Manual install

git clone --depth 1 https://github.com/Adityapal67/context-graph-compressor.git /tmp/context-graph-compressor
mkdir -p ~/.claude/skills
cp -r /tmp/context-graph-compressor/. ~/.claude/skills/context-graph-compressor

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Context Graph Compressor v2

Convert a conversation into a structured, portable JSON state graph — preserving decisions, facts,
problems, goals, code, assumptions, and relationships — so any LLM can resume from the most
important conversational state at minimal token cost.

> **Honest scope:** This preserves the most important state from a conversation. It is not a
> lossless transcript. Some nuance and detail will be lost. The goal is continuity, not reproduction.

---

## Compression Modes

Infer from user intent. Default to `compact`.

| Mode | When | Target tokens |
|---|---|---|
| `compact` | Handoff to new chat, token saving (DEFAULT) | 400–800 |
| `readable` | Human summary, sharing with team, archiving | 1000–2000 |

**`compact` triggers:** "new chat", "hand off", "transfer", "save context", "too many tokens", "start fresh"
**`readable` triggers:** "summarize", "readable", "share", "export", "archive"

---

## Node Types

| Code | Full | Use when |
|---|---|---|
| `F` | Fact | Explicitly stated technical truth — stack, versions, config, schema |
| `D` | Decision | A choice was made, or something was ruled out |
| `P` | Problem | A bug, error, or blocker — with or without resolution |
| `G` | Goal | What the user is trying to build or achieve |
| `C` | Code | A specific snippet, method, annotation, or fix — preserve verbatim |
| `A` | Assumption | Something inferred but never explicitly stated |
| `X` | Context | Open threads, deferred work, unresolved questions |

Read the full skill on GitHub

Details

Skill name: context-graph-compressor
Repository: Adityapal67/context-graph-compressor
License: MIT
Repo updated: 30 May 2026

FAQ

What does the Context Graph Compressor skill do?

How do I install the Context Graph Compressor skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add Adityapal67/context-graph-compressor --skill context-graph-compressor. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Context Graph Compressor popular?

It lives in Adityapal67/context-graph-compressor, which has 35 GitHub stars and 9 forks.

Related skills

Intercom

Trac-Systems/intercom

Skill for autonomous agents. Secure & private P2P messaging (sidechannels), sparse state/data + contracts, and optional value transfer. For a true agentic internet.

AI, LLMs & Agents290

AI Tells

kdgbalmer/ai-tells

Reviews any draft for AI writing tells and produces a clean rewrite. Catches vocabulary fingerprints (delve, tapestry, leverage, robust, multifaceted, navigate, foster), structural fingerprints (negative parallelism like "It's not X, it's Y", bullet-everything, Title Case headings, adjective triads), tone fingerprints ("Great question!", "I'd be happy to help", "In conclusion"), and punctuation fingerprints (em dash overuse, knowledge-cutoff disclaimers). Use this skill whenever the user asks to humanize, de-AI, edit, polish, or sanity-check writing, including phrases like "does this sound lik

AI, LLMs & Agents

R00 Travisvn Awesome Claude Skills Datascience

WireTarantulaKnife/r09-travisvn-awesome-claude-skills-datascience

🤖 Data Science & AI/ML skill suite derived from travisvn/awesome-claude-skills. Data pipelines, model training, evaluation, MLOps and analytical reporting. Provides 10 specialised commands for data-science, machine-learning, analytics workflows.

AI, LLMs & Agents

AI Ad Prompt Guide

creatify-ai/ai-ad-prompt-guide

Battle-tested prompting guide for AI video and image generation in advertising. Covers universal prompting rules, hallucination prevention, the SLCT framework for structured prompts, camera movement vocabulary, UGC formulas, product shot techniques, and the pass³ quality test. Includes model-specific guidance for Sora 2, Veo 3, Kling, Flux, Nano Banana Pro, Seedance, and more. Use when: "write a video prompt", "fix hallucination", "prompt for Sora", "generate B-roll", "UGC prompt", "product shot prompt", "which model should I use", "AI image prompt", "prompt engineering for video", "camera mov

AI, LLMs & Agents

LLM Council

shuntacurosu/llm_council_skill

Orchestrate multiple LLMs as a council, generating collective intelligence through peer review and chairman synthesis

AI, LLMs & Agents

R00 Borghei Claude Skills Datascience

FiendJackdawSilo/r14-borghei-claude-skills-datascience

🤖 Data Science & AI/ML skill suite derived from borghei/Claude-Skills. Data pipelines, model training, evaluation, MLOps and analytical reporting. Provides 10 specialised commands for data-science, machine-learning, analytics workflows.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.