chaudhary-keshav/Warden

Multi-provider event-driven agent that analyzes GitHub PRs using Claude, GPT, or Gemini

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
providerLLM provider to use: claude, openai, or gemini (default: read from warden.toml, falls back to openai)no""
anthropic-api-keyAnthropic API key (sk-ant-...) or OAuth token. Can also be set via ANTHROPIC_API_KEY or CLAUDE_CODE_OAUTH_TOKEN env vars.no
openai-api-keyOpenAI API key. Can also be set via OPENAI_API_KEY env var.no
gemini-api-keyGemini API key. Can also be set via GEMINI_API_KEY env var.no
github-tokenGitHub token for API access (defaults to GITHUB_TOKEN)no${{ github.token }}
config-pathPath to warden.toml config file (relative to repo root)nowarden.toml
fail-onMinimum severity level to fail the action (off, critical, high, medium, low, info). Use "off" to never fail.nohigh
report-onMinimum severity level to show annotations in code review (off, critical, high, medium, low, info). Use "off" to disable reporting.nomedium
max-findingsMaximum number of findings to report (0 for unlimited)no50
request-changesWhether to use REQUEST_CHANGES review event when findings exceed fail-on thresholdnofalse
fail-checkWhether to fail the check run when findings exceed fail-on thresholdnofalse
parallelMaximum number of concurrent trigger executionsno5
namedescription
findings-countTotal number of findings
high-countNumber of high severity findings
summarySummary of the analysis
findings-filePath to structured JSON findings file (always written, use for upload to GCS/S3/etc.)