ScratchLog+: Live Analytics of Learners’ Exercise Progress

[b-fein]

Note

Whisker and the Scratch-GUI are added as Git submodules to this repository. You might want to git clone --recursive to fetch them. Alternatively, you can run git submodule update --init after a regular clone.

---

ScratchLog+

Required Configuration

To connect the embedding connector to an OpenAI-compatible endpoint, create a .env file with the contents

LLM_API_ENDPOINT=http://url.to.my.endpoint/
LLM_API_KEY=if-needed-for-the-provider
# example for https://ollama.com/library/qwen3-embedding,
# other embedding models should work as well
LLM_MODEL=qwen3-embedding:8b

(Optional): Add the pre-seeded database with a demo course

Download the database snapshot and move it to the root of this repository: ase2026_scratchlog-plus_demo.sql.gz

Mount the downloaded file to the database container by configuring it in the docker-compose.yaml file:

services:
  # …
  database:
    # …
    volumes:
      - ase2026_scratchlog-plus_demo.sql.gz:/docker-entrypoint-initdb.d/init.sql.gz
      - scratchlog-db:/var/lib/mysql
    # …

Then, a user admin with password ASE2026tool can be used to log into the ScratchLog user interface.

Running the Application

After this configuration, the application can be started with Docker Compose as described in the readme.

The ScratchLog user interface will be available at http://localhost:8090 and the Scratch user interface at http://localhost:8601.
An initial admin password for ScratchLog will be automatically generated and be printed to the log (docker compose logs scratchlog).

---

See also our extension to Whisker at se2p/whisker#12