noxlie/PR Slop Filter
Detect AI-generated slop PRs before they waste maintainers time
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| threshold | Slop score threshold (0-100) — PRs above this are flagged | no | 70 |
| fail-on-slop | Fail the action if the slop score exceeds the threshold | no | true |
| format | Output format: text, json, or github-annotation | no | github-annotation |
| llm-provider | LLM provider for enhanced detection (openai, anthropic, local) | no | "" |
| llm-api-key | API key for the LLM provider | no | "" |
| llm-model | LLM model name (e.g. gpt-4o-mini, claude-3-haiku) | no | "" |
Outputs
| name | description |
|---|---|
| slop-score | The overall slop score (0-100) |
| verdict | Verdict: human, uncertain, likely-slop, or slop |
| summary | Human-readable summary of the slop analysis |