Agent Daily Review

Helps agents conduct structured end-of-day review, reflection, and documentation. Provides capabilities to scan today's records, categorize activities, perfo...

Security Scan

The skill's code and instructions align with its stated purpose (scanning local workspace notes, categorizing entries, generating a Markdown report, and saving it), but the submitted source was truncated so the assessment is not full-confidence.

Purpose & Capability

The name/description match the provided script: it scans daily memory files and workspace Markdown artifacts, categorizes entries, computes simple analysis, generates a report, and saves/archives it. No unrelated services, credentials, or binaries are requested.

Instruction Scope

SKILL.md and the script instruct the agent to read files from a workspace (default ~/.qclaw/workspace), MEMORY.md, and write reports to reviews/review_YYYY-MM-DD.md and append to MEMORY.md (unless --no-memory). This is expected for the feature, but the skill will access user notes and modify a long-term memory file — users should be aware it reads potentially sensitive local content and (by default) writes/appends to files.

Install Mechanism

No install spec or downloads; the skill is instruction-plus-a-script only. Nothing is fetched from external URLs or installed automatically.

Credentials

The skill declares no required environment variables, credentials, or config paths. The script only uses filesystem paths under the workspace and standard Python libs — no secrets or external services are requested.

Persistence & Privilege

always:false (not force-included). The skill can be invoked autonomously (platform default) which is expected. It does modify files in the user workspace (writes reports and appends to MEMORY.md by default), but it does not request elevated platform privileges or change other skills' configs.

Assessment

This skill appears coherent with its description: it reads local journal files and workspace Markdown files, categorizes entries, produces a Markdown report, and (unless run with --no-memory) appends a summary to your MEMORY.md and saves a report under reviews/. Before installing or scheduling it: 1) review the full script yourself (the provided file content in the prompt was truncated — request the full file if you want a complete audit); 2) if you have sensitive notes, run the script against a copy or use --no-memory to avoid writes; 3) confirm the workspace path (default ~/.qclaw/workspace) and, if you pass a custom -w, point it to a non-sensitive directory; 4) when adding cron, ensure the command path is correct and that you are comfortable with an automated task reading and writing these files; and 5) because this skill runs file I/O on local content, treat it like any tool that can access your notes — inspect the code fully before granting automated execution.

Overview

The Daily Review skill helps agents conduct systematic review and reflection at the end of the day, transforming fragmented daily records into structured growth accumulation.

Core Capabilities:

1. Scan Records - Automatically scan today's memory files, artifacts, and MEMORY.md entries
2. Categorize Activities - Classify activities into: Completed, In Progress, Issues/Blockers, Learning/Growth, Others
3. Reflect and Analyze - Calculate productivity score, identify highlights and challenges, generate improvement suggestions
4. Generate Report - Output structured review report and archive to long-term memory

Use Cases:

• User says "Do my daily review for today"
• User says "Summarize today"
• Cron scheduled task triggers (e.g., daily at 22:00)
• User wants to review work/learning status for a specific day

Workflow

1. Scan Today's Records

Execute scripts/daily_review.py to scan the following:

• memory/YYYY-MM-DD.md - Today's journal entries
• MEMORY.md - Today's entries in long-term memory
• workspace/*.md - Artifact files generated today

2. Categorize Activities

Automatically identify and categorize:

• Completed - Contains keywords like "completed," "done," "resolved," ✅
• In Progress - Contains keywords like "in progress," "working on," 🔄
• Issues/Blockers - Contains keywords like "issue," "blocked," "bug," ❌
• Learning/Growth - Contains keywords like "learned," "researched," "understood"
• Meetings/Communication - Contains keywords like "meeting," "discussed," "sync"

3. Reflect and Analyze

Perform intelligent analysis based on categorization results:

• Productivity Score - Calculate based on record count and artifact count (0-100)
• Today's Highlights - Identify completed important tasks and decisions
• Challenges Encountered - Summarize issues and pending items
• Improvement Suggestions - Generate personalized recommendations based on data

4. Generate Report

Output structured report containing:

• Today's Overview (statistics)
• Completed Tasks List
• In Progress Tasks List
• Issues/Blockers
• Learning/Growth Records
• Highlights Summary
• Reflection and Suggestions
• Tomorrow's Plan Framework

5. Archive to Memory

• Save review report to reviews/review_YYYY-MM-DD.md
• Append review summary to MEMORY.md

Usage

Manual Execution

# Execute today's review
python scripts/daily_review.py

# Specify working directory
python scripts/daily_review.py -w /path/to/workspace

# Specify output file
python scripts/daily_review.py -o /path/to/output.md

# Review specific date
python scripts/daily_review.py -d 2024-01-15

# Do not save to MEMORY.md
python scripts/daily_review.py --no-memory

Use as Module

from scripts.daily_review import DailyReview

review = DailyReview("/path/to/workspace")
report = review.run(save_to_memory=True)
print(report)

Cron Auto-Trigger

Set up automatic daily review at 22:00:

# Add scheduled task using openclaw cron
openclaw cron add --name "daily-review" \
  --schedule "0 22 * * *" \
  --command "python ~/.qclaw/skills/daily-review/scripts/daily_review.py"

Or using cron tool:

{
  "name": "daily-review",
  "schedule": {"kind": "cron", "expr": "0 22 * * *", "tz": "Asia/Shanghai"},
  "payload": {
    "kind": "agentTurn",
    "message": "Please perform today's review using the daily-review skill, scanning today's records and generating a review report."
  },
  "sessionTarget": "isolated"
}

Report Format

Review reports use Markdown format with the following sections:

# Daily Review Report - YYYY-MM-DD

## 📊 Today's Overview
- Date, Record Count, Artifact Count, Productivity Score

## ✅ Completed
- Task List

## 🔄 In Progress
- Pending List

## ⚠️ Issues/Blockers
- Issue List

## 📚 Learning/Growth
- Learning Records

## 🎯 Today's Highlights
- Highlights Summary

## 💭 Reflection and Suggestions
- Improvement Suggestions

## 📝 Tomorrow's Plan
- Plan Framework

Directory Structure

workspace/
├── memory/
│   └── 2024-01-15.md          # Today's journal entries
├── reviews/
│   └── review_2024-01-15.md   # Review report
├── MEMORY.md                   # Long-term memory (review summary appended here)
└── *.md                        # Artifacts generated today

Tips

1. Cultivate Journaling Habit - Record timestamped entries in memory/YYYY-MM-DD.md daily for better review results
2. Use Keywords - Use keywords like "completed," "learning," "encountered issue" when journaling to facilitate auto-categorization
3. Periodic Review - Review weekly/monthly review reports to discover growth trajectory
4. Integrate with Cron - Set up automatic review to ensure daily reflection is never missed

Resources

• scripts/daily_review.py - Core review script
• references/framework.md - Detailed review framework explanation (optional reading)