Analyses
The Analyses page shows optional AI-generated analyses for your project. You only need this page if you enabled an Anthropic or OpenAI provider key; raw logs remain available through the log viewer, REST API, and MCP server either way.
Analysis list
Section titled “Analysis list”Navigate to Analyses in the project sidebar. Each analysis shows:
- Severity badge — high (red), medium (amber), or low (green)
- Title — summary of what the AI provider found
- Timestamp — when the analysis was generated
Skipped analyses
Section titled “Skipped analyses”If your provider’s API key is missing or has been rejected, an analysis is recorded as skipped rather than running. The Analyses page surfaces the past week’s skipped count in an aggregated card so you know whether it’s a one-off or a configuration problem. Add or rotate the key under Settings → AI providers to resume analyses.
Analysis detail
Section titled “Analysis detail”Click an analysis to see the full assessment:
- Root cause — the model’s assessment of what went wrong
- Recommendations — specific steps to resolve the issue
- Related logs — the log entries that triggered this analysis
- Autofix status — if a PR was generated, its current status (open, merged, closed)
Severity levels
Section titled “Severity levels”| Severity | Meaning |
|---|---|
| High | Likely affecting users now. Immediate attention recommended. |
| Medium | Potential issue. Should be investigated soon. |
| Low | Minor concern or informational. Fix when convenient. |