observe

name	observe
description	Force extraction of observations from the current conversation into observational memory
disable-model-invocation	true
argument-hint	[optional: specific topic to focus on]

Observe — Extract Observations from Current Conversation

You are being asked to manually extract observations from the current conversation and save them to the observational memory system.

Steps

1. Determine the memory directory by running:

   python3 -c "import os; print(os.path.expanduser('~/.claude/projects/' + os.getcwd().replace('/','-') + '/memory/'))"
   

2. Read the current observations file at that path (observations.md). If it doesn't exist, you'll create it.

3. Review the current conversation and extract observations following these rules:

Observation Format

Date: YYYY-MM-DD
- HH:MM Observation text
  - HH:MM Sub-detail

What to Extract

• 🔴 High priority: User-stated facts, decisions, preferences, project structure, tech stack
• 🟡 Medium priority: Implementation details, files modified, debugging context
• 🟢 Low priority: Exploratory questions, tentative ideas

Rules

• Distinguish assertions ("User stated X") from questions ("User asked about Y")
• Use absolute dates/times
• Track state changes: "Changed X to Y (previously Z)"
• Preserve specifics: file paths, versions, package names, error messages
• Record outcomes not processes: "Created auth middleware at src/auth.ts" not "Read file, edited file..."
• Capture decisions WITH rationale
• Do NOT repeat anything already in the observations file

4. Append new observations to the observations file. Do not overwrite existing content.

5. Show the user what was observed with a brief summary.

If $ARGUMENTS is provided, focus your observations specifically on that topic.