Quickstart: Bug Prediction

Scan a codebase for patterns that predict bugs before they reach production.

from workflows.bug_predict import BugPredictionWorkflow
from workflows.bug_predict_report import format_bug_predict_report

workflow = BugPredictionWorkflow()
result = workflow.execute(path="src/")
print(format_bug_predict_report(result, {"path": "src/"}))

Expected output:

Bug Prediction Report
Risk Score: 73/100 | Files: 34 | Findings: 8

HIGH (2 findings)
  src/hooks/executor.py:89   dangerous_eval  eval() on user input
  src/plugins/loader.py:142  dangerous_eval  exec() in plugin loader

MEDIUM (3 findings)
  ...

Steps

1. Import the workflow and report formatter

from workflows.bug_predict import BugPredictionWorkflow
from workflows.bug_predict_report import format_bug_predict_report

2. Run the workflow against a path

workflow = BugPredictionWorkflow()
result = workflow.execute(path="src/")

BugPredictionWorkflow coordinates three subagents — pattern-scanner, risk-correlator, and prevention-advisor — and returns a WorkflowResult containing a summary, grouped findings, and prevention suggestions.

3. Format and read the report

print(format_bug_predict_report(result, {"path": "src/"}))

The report groups findings by severity (HIGH, MEDIUM, LOW). Each finding includes a file path, line number, pattern type, and a plain-English description. Start with the HIGH findings — these flag patterns like eval() on user input and broad except clauses that silently swallow errors.

4. Run from the command line

If you prefer a CLI workflow, use the entry point directly:

python -m workflows.bug_predict_report

main() runs the full prediction workflow and prints the formatted report to stdout.

Next: Read How to Run Bug Prediction to learn how to filter by severity, interpret risk scores, and act on findings.

Unresolved references

Auto-generated by attune-author fact-check. Review and either fix the source code, fix this doc, or add an override.