Extracts structured decisions, todos, knowledge, preferences, and risks from AI chat sessions into clean Markdown, JSON, Obsidian, or Notion formats.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install chat-memory-archiver
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install chat-memory-archiver using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Chat Memory Archiver?
The chat-memory-archiver is a powerful tool engineered to automatically parse local AI conversation logs and distill them into highly structured, actionable organizational memory. By scanning raw chat history, it filters out conversational noise to extract concrete engineering decisions, pending task items, newly discovered facts, user configuration preferences, and potential project risks. This ensures valuable insights generated during development sessions are preserved and accessible.
Built to integrate seamlessly with modern personal knowledge management (PKM) ecosystems, this skill prevents context loss across fragmented chat histories. It compiles multi-session logs, merges redundant entries, and formats the output into clean, structured schemas. It serves as an essential utility for developers looking to optimize their workflow using Openclaw Skills to transform conversational data into a reusable knowledge base.
Chat Memory Archiver Use Cases
- Archiving multi-day AI pair-programming sessions into structured markdown files for long-term project documentation.
- Exporting actionable engineering decisions and pending project todos directly into Obsidian vaults or Notion workspaces.
- Generating cross-session dependency graphs to track how technical decisions and topics evolve over the lifecycle of a repository.
- Auditing local session logs for project risks, architectural caveats, and user styling preferences to maintain codebase consistency.
How Chat Memory Archiver Works
- Parse conversation: Reads the specified raw session logs to systematically isolate question-answer pairs, executed tool calls, and major decision points.
- Segment by phase: Labels each conversation segment into contextual phases, specifically categorization under problem, exploration, decision, or action.
- Extract 5 categories: Identifies and extracts key data points into dedicated categories: Decisions, Todos, Knowledge, Preferences, and Risks.
- De-duplicate & merge: Cross-references repeated information discovered across multiple files, fusing duplicates while keeping the latest version accurate.
- Topic tagging: Programmatically appends relevant domain labels and tags like
#pythonor#api-designbased on conversational topics. - Cross-session graph: Establishes a lightweight association graph showing which separate sessions share topics or reference one another.
- Format export: Compiles the final structured data into customized outputs including Markdown, JSON, Obsidian-flavored wiki links, or Notion JSON.
- Summary: Generates a concise, high-level 5-sentence summary of each individual session for rapid scanning and review.
Chat Memory Archiver Setup
Installation
Enable the archiver skill within your workspace configuration or execute the setup via your chosen CLI interface.
Usage Commands
Run the following commands to extract, merge, and export your session histories:
# Extract structured knowledge from specific session log files
session-archiver extract --sessions session-2026-06-01.log session-2026-06-02.log
# Extract all sessions in the directory and export directly to an Obsidian vault folder
session-archiver extract --sessions . --format obsidian --outdir ./vault
# Merge multi-session histories, remove duplicate entries, and output to a JSON file
session-archiver merge --sessions . --dedup --out summary.json
# Generate a cross-session relationship graph in Graphviz DOT format
session-archiver report --sessions . --graph > session-graph.dot
Chat Memory Archiver Data Schema & Taxonomy
The skill processes raw text logs and exports structured data organized into five primary taxonomy categories. The taxonomy structure is mapped out as follows:
| Category | Extracted Metadata & Elements | Target Output Target |
|---|---|---|
| Decisions | Technical choices made, architectural rationale, and alternatives considered | Markdown / Obsidian Wiki |
| Todos | Action items, designated owners, and explicit project deadlines | Notion JSON / Markdown |
| Knowledge | Discovered facts, structural code snippets, relevant URLs, and explanations | Markdown / JSON |
| Preferences | User coding style, distinct terminology, chosen tools, and conventions | JSON Configuration |
| Risks | Security vulnerabilities, known repository issues, and technical caveats | Markdown / JSON |
Session Summaries and Tagging
Each processed session includes metadata block headers featuring:
- Tags: Automatic domain labels (e.g.,
#python,#api-design,#deployment) - Graph Elements: Cross-session association pointers indicating shared conceptual topics
- Executive Summary: A concise 5-sentence summary block defining the session scope
Chat Memory Archiver Advanced Features
- Local-First Parsing: Enhances privacy by parsing all session logs entirely on local infrastructure without transmitting sensitive data to external services.
- Automated PII & Secret Flagging: Automatically scans and flags highly sensitive content such as plaintext passwords or API keys during extraction, requiring explicit user confirmation before inclusion.
- Cross-Session Graphing: Generates visual dependency graphs using the DOT language to map out the relational evolution of architectural decisions across months of conversation.
- Multi-Format Export Engine: Supports simultaneous output compilations tailored for local Markdown environments, Obsidian relational links, and production Notion API schemas.
SKILL.md
Loading
Related Openclaw Skills
Prompt Defender
Prompt Defender is an advanced security scanner that analyzes, scores, and sanitizes AI prompts to block injections, jailbreaks, and sensitive data leaks.
WeChat Digest
Extract, summarize, and synthesize WeChat public account articles into structured knowledge cards.
Codebase Radar
A full-spectrum static analysis tool that scans any codebase for dependency graphs, tech debt hotspots, dead code, and module health scores.
Divorce Litigation Defense Guide
An AI-powered litigation defense guide designed to help defendants navigate the complex legal, financial, and emotional challenges of divorce proceedings.
Trip Planner
An automated travel planning skill that proactive compiles comprehensive, day-by-day itineraries into a polished, mobile-friendly interactive HTML dashboard.
Pet Emotion Recognition
An AI-powered pet emotion analyzer that identifies cat and dog emotions from photos and generates interactive HTML reports with radar charts.
Featured*
