justinkastil/PromptPolygraph

Run the PromptPolygraph evaluation gate (run -> analyze --ci -> report) and emit JUnit/SARIF.

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
configPath to the config.yaml under test.yes
modeCorpus mode passed to `polygraph all` (fixed, varied, adversarial, hybrid). Empty uses the config default.no""
formatComma list of report formats for the human-facing render.nomd,html
baselineBaseline for the regression check (run id, rolling:N, or HEAD). Empty disables it.no""
mockOffline deterministic mode (no API calls). Set "true" for key-free runs.nofalse
python-versionPython version for actions/setup-python.no3.12
extrasOptional pip extras to install with the package, e.g. "llm,pdf". Empty installs core only.no""
sourceInstall source for the package. The PyPI name "promptpolygraph" by default, or "." to install the checked-out repo.nopromptpolygraph
namedescription
verdictGate result, "pass" or "fail".
report-dirDirectory holding the run artifacts (reports, JUnit, SARIF, summary.json).