konflux-ci/Dependency Impact Analysis

Classify dependency PRs by semver bump type, detect risk patterns, and run AI-assisted impact analysis

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
commandCommand to run: classify, analyze, both, or mergeyesboth
pr-numberPull request number (not required for merge command with head-sha)no0
github-tokenGitHub token for API operationsyes${{ github.token }}
api-keyLLM provider API key (required for analyze command)no
llm-providerLLM provider: gemini or claudenogemini
llm-modelLLM model name (provider-dependent default)no
auto-approveApply approved/lgtm labels and formal APPROVE review for eligible low-risk patches and minorsnofalse
auto-mergeMerge eligible PRs after analysis when auto-approve labels are present and CI checks pass (requires auto-approve)nofalse
dry-runSuppress all GitHub API writes; log what would happen insteadnofalse
head-shaHead SHA to find PRs for (used by merge command with check_suite trigger)no
trusted-botsComma-separated list of additional trusted bot logins for supply-chain validation (added to defaults: renovate[bot], red-hat-konflux[bot], dependabot[bot])no
suspicious-pathsComma-separated list of additional suspicious path prefixes for supply-chain validation (added to defaults: .claude/, .vscode/, .github/workflows/, .github/actions/)no
expected-filesComma-separated list of additional expected file patterns for dependency PR scope validation (added to defaults: go.mod, go.sum, vendor/, .tekton/, etc.)no
namedescription
bump-typeDetected semver bump type (major/minor/patch/digest/unknown)
risk-levelAI-assessed risk level (low/medium/high/unknown)
context-jsonPath to the assembled context JSON file