fix: generate browser challenge workspace config

This commit is contained in:
lda
2026-06-15 06:29:25 +07:00 Verified
parent 73dc1d958a
commit 3b19884919
4 changed files with 59 additions and 35 deletions
@@ -19,10 +19,11 @@ uv run wf --config examples/agent_challenges/browser_click_challenge/workspaces/
```
For each local-mode trial, the harness copies `workspace_template/` into
`workspaces/<model>-trial-<n>/` and injects that config path into the prompt.
This builds the configured workflow server in the CLI process for each command
and uses the copied workspace's durable store. It does not reuse in-memory
source sessions across CLI invocations.
`workspaces/<model>-trial-<n>/`, generates a config whose Python source path is
relative to that copied config, and injects the config path into the prompt. This
builds the configured workflow server in the CLI process for each command and
uses the copied workspace's durable store. It does not reuse in-memory source
sessions across CLI invocations.
Use `--start-server` when the trial should exercise the JSON-RPC server path.
With `--start-server`, the harness starts:
@@ -60,10 +61,10 @@ For manual authoring trials, use:
examples/agent_challenges/browser_click_challenge/workspace_template/
```
It contains a local workflow config and prompt template that point at the
browser-click Python source without exposing a generated draft patch answer
file. The harness copies it automatically for normal local-mode trials. For
manual experiments, copy it under:
It contains the prompt template and local store ignore rules without exposing a
generated draft patch answer file. The harness copies it automatically for
normal local-mode trials and writes `wf.config.json` into the copied workspace.
For manual experiments, copy it under:
```text
examples/agent_challenges/browser_click_challenge/workspaces/