# AI use cases for DSH Auto Memory (AskTheWay/dsh-auto-memory)

Repo: [AskTheWay/dsh-auto-memory](https://github.com/AskTheWay/dsh-auto-memory) | Use cases: 4 | Page: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/

A Claude Code-style persistent memory plugin for the DeepSeek Harness (dsh) agent that stores typed memories as plain Markdown files and injects an index into the system prompt.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Stop your dsh agent from forgetting you between sessions: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/persistent-agent-memory-across-sessions/
- Watch the memory write, index, inject and recall cycle offline: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/see-memory-cycle-without-api-key/
- Review and correct the agent's memories as plain files: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/hand-edit-and-audit-memory-files/
- Let the agent file durable facts when a session ends: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/auto-file-facts-at-session-end/

## How to stop your dsh agent from forgetting you between sessions with DSH Auto Memory

Repo: [AskTheWay/dsh-auto-memory](https://github.com/AskTheWay/dsh-auto-memory) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Install the plugin so facts you tell the agent are kept and recalled in new sessions.

**Who it is for:** A developer using DeepSeek Harness daily who repeats the same background and preferences at the start of every new session.

**The problem:** dsh ships with no memory subsystem, so each new session starts without anything the agent learned earlier.

**The outcome:** The agent stores facts you give it as memory files, injects the index into later sessions, and answers questions about you from those memories.

### Steps

1. Run `dsh plugin --profile demo add dsh-auto-memory` to install the plugin from npm.
2. Run `dsh --profile demo` to restart the profile so the plugin loads.
3. In a chat, say something worth keeping, such as "Remember: I'm a Python backend engineer preparing for interviews".
4. Open a brand-new session and ask "what do you know about me?" to confirm the memory is recalled.

### Prompt to paste into your agent

```text
Use the dsh-auto-memory repo to store the fact "<fact to remember>" as a memory for the DeepSeek Harness profile demo. Then start a new session and ask the agent what it knows about the topic. Expected output: the agent quotes the stored fact in the new session. Verify that a Markdown file for the memory exists under $DSH_HOME/memory/.
```

### Caveat

The README describes this as a DeepSeek Harness plugin, so it only helps when you run the dsh agent, not other agent tools.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/persistent-agent-memory-across-sessions/

## How to watch the memory write, index, inject and recall cycle offline with DSH Auto Memory

Repo: [AskTheWay/dsh-auto-memory](https://github.com/AskTheWay/dsh-auto-memory) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Run the bundled demo script to see how memories are written, indexed, injected and forgotten.

**Who it is for:** A developer evaluating the plugin who wants to understand the behavior before connecting a model or API key.

**The problem:** It is hard to judge whether the memory layer does what it claims without first setting up a live agent session.

**The outcome:** You see a full write, index, inject, recall and forget run in your terminal with no API key and no browser.

### Steps

1. Clone the repository from https://github.com/AskTheWay/dsh-auto-memory and open its folder.
2. Run `npm install` in the repository folder to install dependencies. (inferred, check the README)
3. Run `node scripts/demo.mjs` from the repository folder.
4. Read the output and note the injected index section and the forget step.

### Prompt to paste into your agent

```text
Clone the dsh-auto-memory repo into <target-folder>, install its dependencies, and run node scripts/demo.mjs. Expected output: a log of write, index, inject, recall and forget steps with no API key requested. Verify the script exits without error.
```

### Caveat

The demo shows the behavior offline, but the README does not say how the plugin performs inside a live model conversation beyond the stated tests and evals.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/see-memory-cycle-without-api-key/

## How to review and correct the agent's memories as plain files with DSH Auto Memory

Repo: [AskTheWay/dsh-auto-memory](https://github.com/AskTheWay/dsh-auto-memory) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Open the memory folder to read, edit or version the facts your agent has stored.

**Who it is for:** A developer who wants to check what the agent has learned about them and fix wrong facts directly.

**The problem:** Stored memories can be stale or wrong, and you need to see and change them without going through the agent.

**The outcome:** You can read, edit, grep or commit each memory file, and the agent uses the corrected content on its next request.

### Steps

1. Locate the memory folder at $DSH_HOME/memory/, or the memoryDir value if you set one.
2. Open MEMORY.md to see the index and open the linked memory files.
3. Edit or delete the files you want to correct, keeping the YAML frontmatter intact.
4. Commit the folder to git if you want history of memory changes. (inferred, check the README)

### Prompt to paste into your agent

```text
Open the memory folder for the dsh-auto-memory plugin at <memory-folder-path>, list the memory files, and correct the entry named <memory-name>. Expected output: the file reflects the corrected fact. Verify that MEMORY.md still lists that memory and that the next dsh request uses the new content.
```

### Caveat

The README says memories are ordinary files under $DSH_HOME/memory/, but it does not document the exact frontmatter fields, so edit by copying an existing file's structure.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/hand-edit-and-audit-memory-files/

## How to let the agent file durable facts when a session ends with DSH Auto Memory

Repo: [AskTheWay/dsh-auto-memory](https://github.com/AskTheWay/dsh-auto-memory) | Difficulty: Advanced | Time: 30 minutes | Basis: From the README

Turn on auto-summarize so a background model saves new durable facts after each session.

**Who it is for:** A developer whose sessions produce useful decisions and facts that they never write down manually.

**The problem:** Important facts from a session are lost unless the user remembers to ask the agent to store them.

**The outcome:** When a root session ends, a background LLM files up to the configured number of durable new facts as memories.

### Steps

1. Open the profile's cordis.patch.yml file.
2. Add an auto-memory entry with config that sets autoSummarize: true and autoSummarizeMaxMemories: 5.
3. Keep maxBytes and the other values you need, because the config replaces wholesale.
4. Run `dsh --profile demo` to restart the profile, then finish a session and check the memory folder.

### Prompt to paste into your agent

```text
Edit the cordis.patch.yml for the DeepSeek Harness profile demo so the dsh-auto-memory config sets autoSummarize to true and autoSummarizeMaxMemories to 5. Expected output: after a session ends, new memory files appear under the memory folder. Verify that the new files contain durable facts only and that MEMORY.md lists them.
```

### Caveat

Auto-summarize uses a background LLM and is off by default, so review the filed memories because the README does not describe how it judges what is durable.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/asktheway-dsh-auto-memory/auto-file-facts-at-session-end/