Skip to content

CLAUDE.md

Latest commit

 

History

History
41 lines (32 loc) · 1.59 KB

CLAUDE.md

File metadata and controls

41 lines (32 loc) · 1.59 KB
Raw

This repo is to contain various CON demos.

Use repos/ to clone in tools used in demos for AI context.

Each demo should have its own directory.

  • each demo dir should have a venv, which should be used to execute demo
  • each demo dir may provide its own gitignore
  • each demo should have a README with:
    • GIF preview at the top
    • Prerequisites and setup instructions
    • How to generate the recording and GIF

Demo Structure

Demo scripts should:

  • Be executable bash scripts with screencaster say and run commands
  • Activate venv within the script: run "source $SCRIPT_DIR/venv/bin/activate"
  • Run in ~/tmp/ to avoid storing large datasets in the repo
  • Use SCRIPT_DIR for portable paths to scripts/files in the demo directory

Recording and GIF Generation

Generate asciinema recording:

SCREENCAST_HOME=/tmp/demo cast2asciinema script.sh .

Generate GIF with provenance:

datalad run -m "Generate GIF from asciinema recording" \
  --input outputs/script.json \
  --output outputs/script.gif \
  "podman run --rm -v \"\$PWD:/data:Z\" docker.io/kayvan/agg /data/outputs/script.json /data/outputs/script.gif"

Note: Use podman instead of docker to avoid root-owned files. The :Z flag is required for SELinux contexts to allow container write access. Use full image paths like docker.io/kayvan/agg to avoid registry resolution issues.

Top level README should be an index of demos:

  • Title, Short summary, GIF embedded in table.