🛠️ Changes
- Sequential multi-file runs reuse provider and annotation metadata for unchanged test files up to 512 bytes (#1372)
- Parallel reports avoid per-row base64 processes while preserving values and completion order (#1371)
- Data providers avoid per-argument base64 processes while preserving quoting, empty values and parser isolation (#1370)
- JSON reports avoid per-field processes for ordinary filenames, test names and empty messages while preserving escaping (#1369)
- Contributor instructions share task-specific context and model tiers for Claude Code and Codex through a pinned agnostic-ai npm development dependency; shorter workflows align review and release checks with the current code (#1387)
🐛 Bug Fixes
- Releases update and commit project versions in both npm lockfiles and restore them on rollback
- Concurrent provider rows own separate temporary files, even with identical arguments (#1379)
- Synchronous provider rows retain separate results in parallel runs, so later passing rows cannot hide failures (#1377)
- Late timeout signals no longer interrupt
tear_down; the watchdog still bounds hanging cleanup (#1376) - Piped parallel output no longer waits for an inactive spinner's sleep process (#1368)
- Cancellation regression checks distinguish exited test bodies awaiting reaping from running cleanup (#1385)
- Bash compatibility checks scan only shell source files, excluding generated contributor instructions (#1387)
- Skipped macOS and Windows acceptance CI jobs have readable names, with shard labels in test steps
- Linux CI installs the standalone build optimizer required by beta installer acceptance tests (#1383)
👥 Contributors
Checksum
SHA256: 6193370a2164abcc1e64601af461ea811505c6cefdc8fa33e579a07e2d05ffe3
Full Changelog: 0.51.0...0.52.0