@@ -170,20 +170,26 @@ jobs:
170170 - name : Summarize results
171171 if : ${{ success() }}
172172 run : |
173- echo "📊 Evaluation Results" >> $GITHUB_STEP_SUMMARY
174- python -m evaltools summary evals/results --output=markdown >> eval-results.md
175- cat eval-results.md >> $GITHUB_STEP_SUMMARY
173+ echo "## Evaluation results" >> eval-summary.md
174+ python -m evaltools summary evals/results --output=markdown >> eval-summary.md
175+ echo "## Answer differences across runs" >> run-diff.md
176+ python -m evaltools diff evals/results/baseline evals/results/pr${{ github.event.issue.number }} --output=markdown >> run-diff.md
177+ cat eval-summary.md >> $GITHUB_STEP_SUMMARY
178+ cat run-diff.md >> $GITHUB_STEP_SUMMARY
176179
177180 - name : Comment on pull request
178181 uses : actions/github-script@v7
179182 with :
180183 script : |
181184 const fs = require('fs');
182- const summaryPath = "eval-results .md";
185+ const summaryPath = "eval-summary .md";
183186 const summary = fs.readFileSync(summaryPath, 'utf8');
187+ const runId = process.env.GITHUB_RUN_ID;
188+ const repo = process.env.GITHUB_REPOSITORY;
189+ const actionsUrl = `https://github.com/${repo}/actions/runs/${runId}`;
184190 github.rest.issues.createComment({
185191 issue_number: context.issue.number,
186192 owner: context.repo.owner,
187193 repo: context.repo.repo,
188- body: summary
194+ body: `${ summary}\n\n[Check the Actions tab for more details](${actionsUrl}).`
189195 })
0 commit comments