Blame for bundle/README.md

# Geany computer app

Exported by REA's `export-desktop-application` from a capture of the `geany` package
(capture Evidence `ev_4684c2005297be81cf1370fb88326bd891831f5d8ffde813cc0b2e4bca4659a9`). It was checked against agentenv-framework
0.9.1431.

- `envs/geany/` is a computer app on the `ubuntu-24.04` base with the `reference` driver, for
  `amd64` and `arm64`. It installs `geany` from the distribution archive, launches `sh -c exec 'geany' "$HOME"/'Documents/shop/settings.py'` at boot and is ready
  when a window of class `Geany` appears.
- Owned paths: `~/.config/geany`, `~/Documents/shop`.
  `data/get` returns their files and a typed JSON view of each; `data/reset` restores the seed.
- Seed: 6 file(s) from the user's home before the capture, with the captured home path
  respelled in text files.
- `tasks/smoke.json` deploys the app.

Run the task with `agent-env run <this folder>`, or only the deploy with `--task smoke`.

The seed is synthetic test data made for this showcase; it holds no one's real files.

## Task `debug-off`

`tasks/debug-off.json` deploys the app, has the agent work through the computer from the prompt below, and grades
the outcome with `artifacts/debug-off-check/verify.py`, which reads the app's `data/get` views and needs
no model. `evals/geany.toml` names it, so a run of the bundle runs it.

> In the shop's settings.py open in Geany, turn debug mode off and change the port to 8080, then save the file.

Checks, each of which must pass:

- `changed`: `{"kind":"lines_present","file":"Documents/shop/settings.py","lines":["DEBUG = False","PORT = 8080"]}`
- `old-gone`: `{"kind":"lines_absent","file":"Documents/shop/settings.py","lines":["DEBUG = True","PORT = 8000"]}`
- `rest-kept`: `{"kind":"lines_present","file":"Documents/shop/settings.py","lines":["APP_NAME = \"Corner Shop\"","HOST = \"127.0.0.1\"","DATABASE_URL = \"sqlite:///shop.db\"","ITEMS_PER_PAGE = 20","CURRENCY = \"EUR\""]}`

REA kept the task because its checks fail on the seed (`changed`, `old-gone`) and all
pass on the captured end state.

To measure a solve rate, run the task k times (3 by default) with the agent and model to measure, and count the runs
whose summary line says `passed`; a run that says `failed` did not finish and is run again rather than counted:

    agent-env run <this folder> --task debug-off --agent claude-code --model MODEL