# AI use cases for Agent Me (jzjzzzzzzz/agent-me)

Repo: [jzjzzzzzzz/agent-me](https://github.com/jzjzzzzzzz/agent-me) | Use cases: 5 | Page: https://dedupedai.com/use-cases/jzjzzzzzzz-agent-me/

Agent-Me is an open-source Python FastAPI and React system that builds an inspectable AI twin of a person from their reviewable Markdown knowledge, memories, and decisions.

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

- Run the Agent-Me twin locally with Docker Compose: https://dedupedai.com/use-cases/jzjzzzzzzz-agent-me/run-twin-locally-with-docker/
- Inspect why a verified answer was grounded: https://dedupedai.com/use-cases/jzjzzzzzzz-agent-me/inspect-verified-answer-grounding/
- Check a private Markdown folder before startup: https://dedupedai.com/use-cases/jzjzzzzzzz-agent-me/check-private-markdown-before-startup/
- Build your own AI Twin from reviewed Markdown: https://dedupedai.com/use-cases/jzjzzzzzzz-agent-me/build-twin-from-reviewed-markdown/
- Run deterministic evals on the collaboration workflow: https://dedupedai.com/use-cases/jzjzzzzzzz-agent-me/run-deterministic-collaboration-evals/

## How to run the Agent-Me twin locally with Docker Compose with Agent Me

Repo: [jzjzzzzzzz/agent-me](https://github.com/jzjzzzzzzz/agent-me) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Start the web app and API on your machine with the example knowledge and no API key.

**Who it is for:** A developer or curious user who wants to see a personal AI twin running before adding private notes.

**The problem:** Setting up a personal AI system usually needs provider keys and tooling before any behavior can be seen.

**The outcome:** The web application and API run locally using the default extractive and collaboration modes.

### Steps

1. `git clone https://github.com/jzjzzzzzzz/agent-me.git`
2. `cd agent-me`
3. `cp .env.example .env`
4. `docker compose up --build`
5. Open the Web application URL from the Quick start table and ask a question about the example profile.

### Prompt to paste into your agent

```text
Clone the jzjzzzzzzz/agent-me repository and start it from its root with Docker Compose. Report the Web application, Health, and Readiness URLs from the README table, and confirm the readiness check responds. Verify that the default configuration needs no API key before you finish.
```

### Caveat

The default run uses the bundled fictional example profile, so it shows the architecture rather than a real personal twin.

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/jzjzzzzzzz-agent-me/run-twin-locally-with-docker/

## How to inspect why a verified answer was grounded with Agent Me

Repo: [jzjzzzzzzz/agent-me](https://github.com/jzjzzzzzzz/agent-me) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Call the five-stage workflow and read its sources, grounding result, and public trace.

**Who it is for:** An engineer or auditor who needs to see why the twin answered a question the way it did.

**The problem:** A single model response hides how the answer was built, so its evidence is hard to check.

**The outcome:** A JSON response showing the workflow, the exact source excerpts, whether grounding passed, and a per-stage trace.

### Steps

1. `docker compose up --build` from the cloned Agent Me directory
2. `curl http://localhost:8000/api/v1/collaborate -H 'Content-Type: application/json' -d '{"question":"How does the example agent plan a project?","workflow":"verified"}'` (inferred, check the README)
3. Check the grounded field to see whether the implemented evidence policy passed.
4. Read each entry in the trace for role, outcome, safe summary, and metrics.
5. Compare the sources list with the answer to confirm the cited excerpts support it.

### Prompt to paste into your agent

```text
Use the jzjzzzzzzz/agent-me API at localhost:8000 to send the verified collaboration request for <question>. Return the workflow, grounded status, sources, and the trace stage outcomes as a short table. Verify that run_id is present and that each trace stage lists its outcome.
```

### Caveat

The Verifier checks citation paths and output invariants but does not prove that an answer is factually true.

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/jzjzzzzzzz-agent-me/inspect-verified-answer-grounding/

## How to check a private Markdown folder before startup with Agent Me

Repo: [jzjzzzzzzz/agent-me](https://github.com/jzjzzzzzzz/agent-me) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README

Preflight your knowledge directory and get a safe document count and relative file list.

**Who it is for:** A user or automation agent preparing a private notes folder before the twin starts reading it.

**The problem:** A broken or wrong knowledge folder can cause a silent failure, and reading file contents just to check it exposes private text.

**The outcome:** A check result listing the document count and relative paths, with no document contents or absolute host paths.

### Steps

1. `make setup`
2. `make knowledge-check`
3. `.venv/bin/python scripts/check_knowledge.py --knowledge-dir knowledge --json`
4. Compare the reported relative paths with the Markdown files you intended to include.

### Prompt to paste into your agent

```text
In the jzjzzzzzzz/agent-me repository, run the knowledge check against the <knowledge-dir> folder using the JSON option. Return the document count and relative paths only. Verify that the output contains no document text and no absolute host paths.
```

### Caveat

The check reports only file count and paths, so you still need to read the Markdown yourself before approving it.

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/jzjzzzzzzz-agent-me/check-private-markdown-before-startup/

## How to build your own AI Twin from reviewed Markdown with Agent Me

Repo: [jzjzzzzzzz/agent-me](https://github.com/jzjzzzzzzz/agent-me) | Difficulty: Advanced | Time: 2 hours | Basis: From the README

Replace the example profile with your approved notes, set the app identity, and run the checks.

**Who it is for:** A developer who wants a personal twin grounded in their own reviewed notes and is ready to maintain evaluation cases.

**The problem:** The example profile is fictional, so the system does not yet represent a real person.

**The outcome:** A private configuration with your own Markdown corpus that passes lint, tests, docs checks, and evaluation.

### Steps

1. Replace knowledge/example-profile.md with reviewed Markdown you are allowed to use.
2. Set APP_NAME and APP_DESCRIPTION in your private .env file.
3. `make knowledge-check`
4. Add supported, unsupported, adversarial, and boundary evaluation cases under course/fixtures.
5. `make lint test docs evaluate`

### Prompt to paste into your agent

```text
Update the jzjzzzzzzz/agent-me checkout so the knowledge folder contains only the reviewed Markdown files I approve, then run the knowledge check. Report the document count and the output of make lint test docs evaluate. Verify that no file outside the approved list was added.
```

### Caveat

The README says to configure an OpenAI-compatible provider only when its data-processing boundary is acceptable, so review that before adding private notes.

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/jzjzzzzzzz-agent-me/build-twin-from-reviewed-markdown/

## How to run deterministic evals on the collaboration workflow with Agent Me

Repo: [jzjzzzzzzz/agent-me](https://github.com/jzjzzzzzzz/agent-me) | Difficulty: Intermediate | Time: 60 minutes | Basis: From the README

Run the versioned fixtures and see the pass count for the verified workflow.

**Who it is for:** A maintainer or evaluator who needs repeatable results before trusting changes to the role pipeline.

**The problem:** Behavior of a multi-role AI pipeline can drift without anyone noticing unless it is measured the same way each time.

**The outcome:** A deterministic report of supported, unsupported, adversarial, and boundary cases for the verified workflow.

### Steps

1. `make setup`
2. `make evaluate`
3. `.venv/bin/python scripts/evaluate_collaboration.py --workflow verified --json`
4. Confirm the default evaluator reports COLLABORATION_EVAL 4/4 passed.

### Prompt to paste into your agent

```text
In the jzjzzzzzzz/agent-me repository, run the collaboration evaluator with the verified workflow in JSON mode. Return the pass count and any failed case IDs. Verify that rerunning the same command on unchanged code gives the same result.
```

### Caveat

The fixtures are small and fictional, so a passing score shows that the checks run, not that the twin represents a real person well.

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/jzjzzzzzzz-agent-me/run-deterministic-collaboration-evals/