levantar-ai/loadr load test
Run a loadr load test in CI. Installs loadr, runs a test plan, writes a JUnit XML report + JSON summary, and fails the job when a threshold is breached. Optionally diffs the run against a baseline summary and comments the comparison on the pull request.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| plan | Path to the loadr test plan (YAML). | yes | — |
| version | loadr version to install (a tag like "v1.21.5", or "latest"). | no | latest |
| bin | Path to a preinstalled loadr binary. When set, the download is skipped and this binary is used (handy for self-hosted runners that already ship loadr). | no | "" |
| junit | Path to write the JUnit XML report. | no | loadr-junit.xml |
| summary | Path to write the JSON summary export. | no | loadr-summary.json |
| args | Extra arguments passed verbatim to `loadr run`. | no | "" |
| fail-on-threshold | When 'true' (default) a breached threshold fails the step. Set 'false' to record results without failing the job. | no | true |
| baseline-summary | Path to a baseline JSON summary from a previous run. When set, the action runs `loadr compare` between it and this run's summary. | no | "" |
| compare-markdown | Path to write the baseline comparison as markdown. | no | loadr-compare.md |
| max-regression | Whitespace-separated regression tolerances, passed to `loadr compare` as repeated `--max-regression` flags (e.g. 'p95=10% error_rate=0.5'). | no | "" |
| fail-on-regression | When 'true' (default) a regression beyond tolerance fails the job — after the PR comment is posted. Set 'false' to record the comparison without failing. | no | true |
| comment-pr | When 'true' (default) the comparison markdown is upserted as a single PR comment. Only acts on pull_request events. | no | true |
Outputs
| name | description |
|---|---|
| passed | 'true' when every threshold passed and the run was not aborted. |
| exit-code | The exit code from `loadr run`. |
| junit | Path to the JUnit XML report. |
| summary | Path to the JSON summary. |
| regression | 'true' when the baseline comparison found a regression beyond tolerance, 'false' when it ran clean; empty when no baseline-summary was given. |
| comparison | Path to the comparison markdown (when baseline-summary is set). |