Memory Management

skills.sh install memory-management
``` Or, add to your `skills.yaml`: ```yaml
- name: memory-management source: https://github.com/aaron-he-zhu/seo-geo-claude-skills/tree/main/cross-cutting/memory-management
``` ### Usage Invoke the skill with one of its supported triggers, or use the argument-hint for specific actions: ```text
[review|archive|cleanup]
``` **Example:** ```text
Claude, please review my project memory.
``` Or programmatically via the Claude Code API: ```python
from claude_skills import MemoryManagement mm = MemoryManagement
mm.review # Review all active campaign memory
mm.archive # Move old items to archive tier
mm.cleanup # Purge stale or redundant context
``` ## Key Features **1. Automatic Tiered Storage** Memory Management segments project data into three tiers: - **Hot-list:** Recently updated or high-priority context (e.g., new keyword findings, urgent campaign tasks).
- **Active Work:** Ongoing campaign details, discussions, and pending items.
- **Archive:** Completed, obsolete, or less relevant history for reference. **2. Flexible Triggers and Natural Language Control** The skill recognizes a wide range of English and multilingual cues, including: - “remember project context”
- “save SEO data”
- “what did we decide last time”
- “” (Chinese) **3. Integration-Ready** Memory Management works standalone or as part of a broader Claude Code skillset. Optional MCP network access allows for direct integration with third-party SEO tools, making it possible to automatically ingest data from analytics platforms or keyword trackers. **4. Efficient Memory Review and Cleanup** Effortlessly review current project status, clean up outdated memory, and archive items with simple prompts or API calls:

```python
mm.cleanup
mm.archive
mm.review
``` **5. Multilingual Support** The skill recognizes and responds to triggers in English, Chinese, Japanese, Korean, and Spanish (memoria-proyecto), making it adaptable for global teams. ## Best Practices - **Define Clear Project Boundaries:** Clearly label campaign context when saving to minimize ambiguity in future sessions.
- **Regularly Review and Archive:** Use the `review` and `archive` actions to keep active memory relevant and concise.
- **Leverage Hot-list for Urgent Items:** Use the hot-list tier for time-sensitive discussions, tasks, or findings.
- **Automate Data Ingestion:** Integrate with MCP or other SEO tools to avoid manual data entry and ensure data freshness.
- **Use Natural Language:** Take advantage of the skill’s flexible, multilingual triggers for seamless workflow integration. **Example Workflow:**

```text
Claude, save my keyword research findings to the project memory.
Claude, what did we decide last time about link building?
Claude, archive completed backlink audit tasks.
``` ## Important Notes - **Privacy and Security:** Ensure that sensitive campaign information is managed according to your organization’s data policies when using persistent memory features.
- **Skill Compatibility:** Memory Management requires Claude Code ≥1.0. No additional system packages are needed.
- **Performance:** Excessive memory accumulation may impact performance; regularly archive and clean up context to maintain efficiency.
- **Optional Integrations:** MCP network access is not required but offers advanced integration capabilities for SEO data pipelines.
- **Geo-Relevance:** The skill is primarily designed for SEO/GEO campaign tracking, with low geo-specific limitations. Memory Management is an essential productivity tool for teams seeking consistent, reliable campaign context across Claude sessions. By leveraging its tiered memory architecture and seamless integration, users can maintain project momentum and organizational clarity across even the most complex SEO or GEO initiatives.