Compare runs
B is compared with the latest successful run of the previous version.
11 steps · 13.4s · ≈$0.048
14 steps · 19.2s · ≈$0.086
steps 11 → 14 · cost +80% · duration +44% · outcome succeeded
What changed between them
| Agent version | 1.4.0 | 1.5.0 |
| Prompt version | v7 | v8 |
Step by step
= same (4)≠ changed (5)+ only in B (4)− only in A (1)| A · run_8b4a65d4 | B · run_3ee8bfeb | |
|---|---|---|
| = | #2 claude-sonnet-4-5 | #2 claude-sonnet-4-5 |
| ≠ | #3 sentry.getIssue | #3 sentry.getIssueargs changed |
| = | #4 claude-sonnet-4-5 | #4 claude-sonnet-4-5 |
| ≠ | #5 gitlab.searchCode | #5 gitlab.searchCodeargs changed |
| + | — | #6 claude-sonnet-4-5 |
| + | — | #7 gitlab.searchCode |
| + | — | #8 claude-sonnet-4-5 |
| + | — | #9 gitlab.searchCode |
| ≠ | #6 gitlab.readFile | #10 gitlab.readFileargs changed |
| = | #7 claude-sonnet-4-5 | #11 claude-sonnet-4-5 |
| ≠ | #8 jira.createIssue429 Too Many Requests | #12 jira.createIssue429 Too Many Requestsargs changed |
| ≠ | #9 jira.createIssue429 Too Many Requests | #13 jira.createIssueargs, outcome changed |
| − | #10 jira.createIssue | — |
| = | #11 slack.postMessage | #14 slack.postMessage |
System spans are left out; steps are aligned on the tool called (or LLM call), then compared on arguments and outcome.