记忆

Session Archiver Pro

试用

Extract decisions, todos, knowledge, preferences, and risks from AI chat sessions into structured memory. Outputs Markdown, JSON, or Obsidian notes.

它能做什么

Extract decisions, todos, knowledge, preferences, and risks from AI chat sessions into structured memory. Outputs Markdown, JSON, or Obsidian notes.

技能文档

Session Archiver Pro

Extract structured memory from AI chat sessions. Transform raw conversation logs into organized knowledge: decisions, todos, knowledge points, user preferences, and risks.

Why Session Archiver Pro?

  • Persist what matters: No more scrolling through endless chat logs to find that one decision.
  • Cross-session awareness: Link related information across different conversations.
  • Knowledge management ready: Export to Markdown, JSON, or Obsidian — plug into your note-taking workflow.
  • Agent memory injection: Outputs are formatted for direct injection into long-term agent memory.

Workflow (9 Steps)

Input: One or more session logs (plain text / .log / .json / clipboard)
↓
[1] Parse dialogue structure: user messages, assistant replies, tool calls, decision points
[2] Segment by conversation arc: problem definition → exploration → decision → action
[3] Extract 5 structured categories:
    📌 Key Decisions — explicit choices/conclusions
    ✅ Action Items — who needs to do what by when
    📚 Knowledge Points — reusable facts or insights
    ⭐ User Preferences — expressed likes/dislikes, style, standards
    ⚠️ Risks & Issues — potential problems identified
[4] Deduplicate & merge: consolidate identical or related items across sessions
[5] Generate topic tags: auto-classify each session with hierarchical tags
[6] Build cross-session graph: link related items ("this decision relates to last week's todo")
[7] Export in chosen format: Markdown report / JSON / Obsidian-compatible notes
[8] Optional: format for agent long-term memory injection
[9] Produce executive summary: 5-sentence overview for quick recall
↓
Output: Structured memory report + tag index + cross-session links + summary

When to Use

  • After a long planning session with an AI assistant
  • Before starting a new conversation on a related topic
  • During weekly review of AI-assisted work
  • When onboarding a new team member to an existing project
  • Before an AI agent's context window resets

Sample Prompts

Sample 1: Daily session summary

"Summarize today's 3 chat sessions about the SaaS pricing model. Extract decisions, action items, and open questions."

Sample 2: Cross-session knowledge mining

"I have 12 sessions from the past week on our product launch. Find all decision points and create a timeline."

Sample 3: Preference extraction

"From all my conversations with the code assistant, extract my coding style preferences (linting, testing, naming conventions)."

Sample 4: Risk identification

"Scan these 5 design review sessions and flag any risk signals or tensions I should revisit before finalizing."

Sample 5: Agent memory prep

"Archive this session about database migration planning. Output in agent-memory-injection format for my AI assistant's long-term memory."

First-Success Path

1. Run: python3 scripts/archiver.py --file sample_session.log
2. Review the structured output in stdout (Markdown by default)
3. Use --format json for programmatic consumption
4. Use --format obsidian for note-taking integration
5. Use --dir ./sessions/ for batch processing multiple sessions

Output Examples

Markdown Report (Decision-Centric)

# Session Memory Report — SaaS Pricing (3 sessions)

## 📌 Key Decisions (4)
| # | Decision | Session | Impact |
|---|----------|---------|--------|
| 1 | Usage-based pricing model | Session-1 | Revenue model shift |
| 2 | Target audience: SMB (10-50 seats) | Session-1 | Product positioning |
| 3 | Free tier limit: 1K MAU | Session-2 | User acquisition funnel |
| 4 | No annual-only billing | Session-3 | Cash flow consideration |

## ✅ Action Items (6)
- [ ] Validate competitive pricing (John, by Fri)
- [ ] Conduct 5 customer interviews (Team, next week)
- [ ] Design pricing page mockup (Design, Mon)

## 📚 Knowledge Points (8)
- Industry conversion rate: 3-5% free→paid
- AWS pricing uses consumption + reservation hybrid
- Competitors all offer annual discount (15-20%)

## ⭐ User Preferences (3)
- Prefers clean, simple pricing (no feature tiers overload)
- Dislikes hidden fees or "starting at" bait-and-switch
- Wants transparent cost calculation upfront

## ⚠️ Risks (2)
- Free tier compute costs may exceed budget at scale
- Minimum commit could scare SMB segment

JSON Output Snippet

{
  "sessions": ["session-1", "session-2", "session-3"],
  "decisions": [{"text": "Usage-based pricing model","session": "session-1","tags": ["pricing","revenue-model"]}],
  "knowledge_points": [{"text": "Industry conversion rate: 3-5% free→paid","source": "session-2"}]
}

CLI Usage

python3 scripts/archiver.py --help
  --file PATH         Single session log file
  --dir PATH          Directory of session logs
  --format FORMAT     Output format: markdown (default), json, obsidian, memory-inject
  --tags              Enable auto-tagging (default: on)
  --graph             Enable cross-session linking (default: on)

Tags

session-archiver-pro, memory, session, archive, extract, knowledge-management, obsidian, chat-summary

相关技能

Extract decisions, todos, knowledge, preferences, and risks from AI chat sessions into structured memory

5 次安装

Unlimited organized memory for your AI agent. Store, search, and organize projects, contacts, decisions, and knowledge across categories. Never lose context...

Turn raw meeting notes or transcripts into structured action: decisions log, action items with owner and deadline, open questions, and a distributable summary email. Extracts who-committed-to-what with date parsing (natural language dates resolved against the meeting date), deduplicates, tracks carryover between meetings, and generates per-owner task lists. Use when the user has meeting notes/transcripts and needs minutes, action items, or a follow-up summary.

1 次安装

Durable memory for projects and research

1 次安装

在本地 Markdown 笔记库(Obsidian/Logseq 风格)中读写、检索并关联笔记,写入需用户授权,追加内容附带署名。

22 次安装