# Weekly Summary Playbook

This playbook helps you generate concise weekly summaries, ideal for reporting to stakeholders or for personal reflection. It leverages AI to distill key information from your session memory, task lists, and other relevant inputs.

## 1. Prompt Pack

Use these prompts to guide your AI in generating a weekly summary. Adapt them based on the specific information you want to highlight.

### Core Weekly Summary Prompt

```
"You are an expert assistant generating a weekly summary. Review the provided context (task history, memory files, project updates) and synthesize a concise summary for the past week. Focus on:
- **Key Accomplishments:** What important tasks were completed?
- **Significant Decisions:** Any major choices made or approved?
- **Blockers/Challenges:** What issues arose and how were they addressed (or are being addressed)?
- **Upcoming Priorities:** What are the main focuses for the next week?
- **Lessons Learned:** Any insights gained or process improvements identified?

Keep it direct, professional, and under 300 words. Format as bullet points under clear headings."
```

### Prompt Variations:

*   **Stakeholder-Focused Summary:**
    ```
    "Generate a weekly summary for project stakeholders. Emphasize progress against goals, critical decisions, and any high-level risks/mitigations. Keep technical details minimal. Focus on business impact."
    ```
*   **Detailed Technical Summary:**
    ```
    "Provide a detailed technical weekly summary for the engineering team. Include specific feature completions, bug fixes, architecture discussions, and any relevant technical challenges or breakthroughs. Refer to ticket numbers where possible."
    ```
*   **Personal Reflection Summary:**
    ```
    "Create a personal weekly reflection. What did I achieve? What did I struggle with? What am I proud of? What do I want to improve next week? Focus on personal growth and learning."
    ```

## 2. Setup Guide

To set up and use this playbook:

1.  **Save the playbook:** Save this content as `weekly-summary-playbook.md` in your OpenClaw workspace.
2.  **Create a configuration file:** Create a `weekly-summary-cron.json` (or `.yaml`) file in your workspace with the content provided in the "Config Snippet" section below. This will configure a cron job to run the summary generation automatically.
3.  **Ensure Context:** For the AI to generate a meaningful summary, ensure your OpenClaw agent has access to its session history, daily memory files (`memory/YYYY-MM-DD.md`), and any project-specific updates (e.g., `shared-context/agent-outputs/`). The prompt will instruct the AI to review this available context.
4.  **Activate Cron Job:** Once the `weekly-summary-cron.json` file is in your workspace, OpenClaw should automatically detect and activate the cron job. You can verify its status using `openclaw cron list`.
5.  **Manual Execution:** To run the summary manually at any time, you can instruct your agent to execute the core prompt with relevant context.

## 3. Config Snippet

This JSON snippet defines a cron job that will trigger the agent to generate a weekly summary. Save this as `weekly-summary-cron.json` in your workspace.

```json
{
  "cron": [
    {
      "name": "Weekly Summary Generation",
      "schedule": "0 9 * * 1",
      "task": "Generate a concise weekly summary based on my recent activity and memory files, following the 'Core Weekly Summary Prompt' from the 'Weekly Summary Playbook'. Save the output to memory/weekly-summary-{YYYY-MM-DD}.md.",
      "agent": "cmo",
      "model": "gemini-flash"
    }
  ]
}
```

**Explanation of fields:**
*   `name`: A descriptive name for the cron job.
*   `schedule`: A standard cron string (`0 9 * * 1` means "At 09:00 on Monday").
*   `task`: The prompt that will be sent to the agent when the cron job runs. It references the `Core Weekly Summary Prompt` from this playbook.
*   `agent`: The agent assigned to run this cron job (`cmo` in this case).
*   `model`: The model to use for this specific cron task.

## 4. Permissions Checklist

To ensure this playbook runs successfully, the agent requires the following:

*   **Read Access:**
    *   `memory/YYYY-MM-DD.md`: To access daily logs and context.
    *   `MEMORY.md`: For long-term memory and curated insights (if applicable to the agent's context).
    *   `shared-context/agent-outputs/`: To read outputs from other agents (if cross-agent context is desired).
    *   OpenClaw session history: The agent inherently has access to its own session history.
*   **Write Access:**
    *   `memory/weekly-summary-{YYYY-MM-DD}.md`: To save the generated weekly summary.
    *   `weekly-summary-cron.json`: To create/modify the cron configuration file.
*   **Execution Permissions:**
    *   Ability to process and respond to cron-triggered tasks.

By following these guidelines, your agent will be equipped to consistently generate valuable weekly summaries.
## Safety Check (Required)

Before production use:

- Review prompts for prompt-injection risk.
- Remove or rewrite any instructions requesting unnecessary secrets, unrelated external actions, or policy bypasses.
- Test with non-sensitive data first.
- Keep manual approval for outbound or destructive actions.