How do I install the Memory Consolidator skill?

Run npx skillmds@latest add officebeats/memory-consolidator in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Memory Consolidator safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: executes scripts. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Memory Consolidator?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Memory Consolidator free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Memory Consolidator?

officebeats (@officebeats) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Memory Consolidator by officebeats — Agent Skill

Name: Memory Consolidator
Rating: 4.2 (5 reviews)
Author: officebeats

File contents

Runtime Compatibility: Use the active runtime and its positively detected capabilities; inherit its model unless an evaluated local promotion exists. Write Preference: For cloud-based paths (iCloud/CloudDocs), utilize shell-based write operations (cat >) to ensure atomic persistence and avoid tool-cancellation.

memory-consolidator

Goal

The memory-consolidator mimics the human brain's "sleep state." It reads unconsolidated memories (recent transcripts, tasks, issues) from the prior 24-hours to 1-week, explicitly looking for connections, risks, or systemic issues that cross-cut individual files.

Inputs

Provided as target file paths or directory scope by the caller. E.g. 1. Company/STRATEGY.md, 3. Meetings/transcripts/, etc.
Default to analyzing modifying files over the prior 48 hours within 2. Products/ and 3. Meetings/.

Behaviors

Extraction: Identify core entities, decisions made, risks raised, and blockers.
Correlation: Find where concepts overlap (e.g. Issue A in bug-tracker is secretly related to Decision B in the product transcript).
Draft Insights: Derive the top 1-3 highest leverage insights. An insight must not just summarize; it must synthesize (X + Y = Z).

Template Output Rules

You will generate an output block for the insights that conforms to the following structure:

### 🧠 Deep Synthesis (YYYY-MM-DD)
**Scope:** [Brief note on what was analyzed, e.g. Transcripts & Bug Trackers]

- **Insight 1 [Category]:** A bold claim synthesizing multiple data points.
  * *Evidence:* [File Reference X] + [File Reference Y]
  * *Recommendation:* What the user should do about this today.

- **Insight 2 [Category]:** Another bold synthesizer.
  * *Evidence:* [File Reference X] + [File Reference Y]
  * *Recommendation:* Actionable next step.

File Targets

STRATEGIC_INSIGHTS: Append the full template string above to 5. Trackers/STRATEGIC_INSIGHTS.md.
STATUS.md: Take ONLY the bold claims for each insight and prepend them as a short bulleted list strictly underneath the "## 🧠 Latest Strategic Insights" header in the root STATUS.md file.

---name: memory-consolidatordescription: Consolidates memories and generates cross-cutting strategic insights from disparate files.---> **Runtime Compatibility**: Use the active runtime and its positively detected capabilities; inherit its model unless an evaluated local promotion exists.> **Write Preference**: For cloud-based paths (iCloud/CloudDocs), utilize shell-based write operations (`cat >`) to ensure atomic persistence and avoid tool-cancellation.# memory-consolidator## GoalThe memory-consolidator mimics the human brain's "sleep state." It reads unconsolidated memories (recent transcripts, tasks, issues) from the prior 24-hours to 1-week, explicitly looking for connections, risks, or systemic issues that cross-cut individual files. ## Inputs- Provided as target file paths or directory scope by the caller. E.g. `1. Company/STRATEGY.md`, `3. Meetings/transcripts/`, etc.- Default to analyzing modifying files over the prior 48 hours within `2. Products/` and `3. Meetings/`.## Behaviors1. **Extraction**: Identify core entities, decisions made, risks raised, and blockers.2. **Correlation**: Find where concepts overlap (e.g. Issue A in bug-tracker is secretly related to Decision B in the product transcript).3. **Draft Insights**: Derive the **top 1-3 highest leverage insights**. An insight must not just summarize; it must synthesize (X + Y = Z).## Template Output RulesYou will generate an output block for the insights that conforms to the following structure:```markdown### 🧠 Deep Synthesis (YYYY-MM-DD)**Scope:** [Brief note on what was analyzed, e.g. Transcripts & Bug Trackers]- **Insight 1 [Category]:** A bold claim synthesizing multiple data points. * *Evidence:* [File Reference X] + [File Reference Y] * *Recommendation:* What the user should do about this today.- **Insight 2 [Category]:** Another bold synthesizer. * *Evidence:* [File Reference X] + [File Reference Y] * *Recommendation:* Actionable next step.```## File Targets1. **STRATEGIC_INSIGHTS**: Append the full template string above to `5. Trackers/STRATEGIC_INSIGHTS.md`.2. **STATUS.md**: Take ONLY the bold claims for each insight and prepend them as a short bulleted list strictly underneath the "## 🧠 Latest Strategic Insights" header in the root `STATUS.md` file.

Frequently asked questions

Memory Consolidator