fix: polish azd eval output - dataset path, Rows section, Foundry link, terser logs#297
Merged
Merged
Conversation
…k, terser logs Four user-visible issues with `execution: azd` runs: 1. `report.md` shipped an empty `Dataset:` line because the eval.yaml parser only recognized `dataset_reference:` and `azd ai agent eval init` actually emits `dataset_file:`. EvalRecipe now accepts `dataset_file:` and `_recipe_dataset_path` prefers it. 2. `report.md` shipped a header-only `## Rows` table on every azd run (azd does not expose per-row metrics through `agentops eval run`). The reporter now omits the row sections when `result.rows` is empty and instead emits a `## Per-row breakdown` callout linking to the Foundry run. 3. The CLI did not surface the Foundry run URL. `azd_runner.normalize_to_results` now captures `report_url` from the azd payload and the CLI prints a `Foundry run:` line next to the results paths. 4. `Running azd backend: azd --no-prompt ai agent eval run --config <long absolute path> --output json` was unreadable; replaced with `Running azd backend: azd ai agent eval run` (full command stays in the failure debug logs added in 0.3.18). The `delegating to azd ai agent eval` startup line also uses a workspace-relative recipe path. 921 unit tests pass. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Four user-visible papercuts in execution: azd runs:
report.md showed Dataset: with an empty value. eval.yaml parser only recognized dataset_reference but azd ai agent eval init emits dataset_file. EvalRecipe now accepts dataset_file and _recipe_dataset_path prefers it.
report.md always shipped a header-only ## Rows table (azd does not expose per-row metrics through agentops eval run). The reporter now omits the row sections when rows is empty and emits a ## Per-row breakdown callout linking to the Foundry run.
CLI did not surface the Foundry run URL. normalize_to_results captures report_url from the azd payload; CLI prints a Foundry run: line next to results.json / report.md / latest/.
The Running azd backend log line repeated the full command with absolute Windows paths. Replaced with the short Running azd backend: azd ai agent eval run (full command stays in the failure debug logs added in 0.3.18). The delegating to azd ai agent eval startup line also uses a workspace-relative recipe path.
921 unit tests pass.