Behavior Tracker

Automatically record and analyze user behavior patterns from conversations

Security Scan

The skill's code and instructions are consistent with its stated purpose: it reads local conversation memory files, extracts keywords, and writes local behavior reports and aggregates — no external network calls or unrelated credential access detected.

Purpose & Capability

Name/description (behavior tracking) align with the code: analyzer/recorder/heartbeat read conversation memory files and update local behavior-patterns.json and reports. All requested resources (local workspace/memory files) are reasonable for this purpose.

Instruction Scope

SKILL.md instructs running analyzer, recorder, and heartbeat scripts and recommends a cron entry. The scripts only read/write files under the OpenClaw workspace memory directory, extract keywords, and generate reports. They do not read unrelated system files or transmit data externally.

Install Mechanism

No install spec; the skill is instruction-only with small Python scripts using only the standard library. Nothing is downloaded or installed by the skill.

Credentials

The skill declares no required env vars or credentials (good). The code does read optional environment variables OPENCLAW_DIR and OPENCLAW_WORKSPACE (with safe defaults) but these were not listed in metadata — this is benign but could be declared explicitly. No secrets, network endpoints, or external credentials are requested.

Persistence & Privilege

always:false and the skill does not request elevated privileges. It writes persistent local files (behavior-patterns.json, behavior-report.md, behavior-last-check.json) into the user's OpenClaw workspace memory directory — expected given its purpose; users should review/clean these files if they contain sensitive content.

Assessment

This skill operates on local conversation memory files and stores aggregated behavior data locally — there are no external network calls or credential requests. Before installing: (1) confirm the workspace path (~/.openclaw/workspace/memory by default) contains only data you want analyzed, (2) be aware it will write behavior-patterns.json, behavior-report.md, and a last-check file into that directory, (3) if you want different storage, set OPENCLAW_DIR or OPENCLAW_WORKSPACE or run the scripts from a controlled location, and (4) if you schedule cron jobs, verify the cron command and environment to avoid accidental exposure. If you need the skill to declare the optional env vars explicitly, request that from the author.

Automatically record and analyze user behavior patterns

Features

1. Proactive Recording - Extract key info after each conversation
2. Periodic Analysis - Daily/weekly auto analyze behavior patterns
3. Pattern Recognition - Discover interests, active hours, learning habits
4. Report Generation - Generate visual reports

Trigger Conditions

• Auto call after each important conversation
• Cron: Daily 18:00 analysis
• User request: "analyze my behavior"

Analysis Dimensions

Dimension	Description
topic	Recently discussed tech/projects
active_hours	High-frequency conversation time periods
communication_style	Task-based/discussion-based/Q&A
learning_mode	Theory-oriented/practice-oriented
project_focus	Long-term projects of interest

File Structure

behavior-tracker/
├── SKILL.md              # This file
├── scripts/
│   ├── analyzer.py       # Core analysis logic
│   ├── recorder.py       # Manual conversation recorder
│   └── heartbeat_tracker.py  # Lightweight heartbeat tracker
└── references/
    └── .gitkeep

Usage

# Manual analysis run
python3 scripts/analyzer.py

# Record current conversation
python3 scripts/recorder.py --topic "AI Agent" --project "The Machine"

# Run lightweight tracker (for heartbeat)
python3 scripts/heartbeat_tracker.py

Cron Configuration

# Auto analysis daily at 18:00
0 18 * * * python3 ~/.../behavior-tracker/scripts/analyzer.py

Dependencies

• Python 3.10+
• No external dependencies (pure stdlib)