Use Case: Automated Code Review in CI/CD

The Problem

Manual code review is slow. You want AI to catch common issues -- security vulnerabilities, performance problems, style violations -- before human reviewers even look at the PR.

Solution

Use non-interactive mode with structured output to get machine-readable results.

Basic: Review from Stdin

# Feed a diff to Octomind and get a review
git diff main..HEAD | octomind run developer \
  --format plain <<< "Review this diff for security issues, performance problems, and bugs. Be specific about file and line numbers."

Structured: JSON Output for Pipeline Decisions

Prompt the AI to return JSON directly:

#!/bin/bash
# ci-review.sh

diff=$(git diff main..HEAD)
result=$(echo "Review this diff for issues. Respond in JSON with: {summary, issues: [{file, line, severity, description}], approval}.\n\n$diff" | \
  octomind run developer \
  --model openai:gpt-4o \
  --format plain)

approval=$(echo "$result" | jq -r '.approval')
errors=$(echo "$result" | jq '[.issues[] | select(.severity == "error")] | length')

echo "Review: $approval ($errors errors)"

if [ "$approval" = "request_changes" ] || [ "$errors" -gt 0 ]; then
  echo "$result" | jq '.issues[]'
  exit 1
fi

# .github/workflows/ai-review.yml
name: AI Code Review
on: [pull_request]

jobs:
  review:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - name: Install Octomind
        run: curl -fsSL https://raw.githubusercontent.com/muvon/octomind/master/install.sh | bash

      - name: Run AI Review
        env:
          OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
        run: |
          diff=$(git diff origin/main..HEAD)
          result=$(echo "$diff" | octomind run developer \
            --format plain <<< "Review for security, performance, bugs")
          echo "$result" >> $GITHUB_STEP_SUMMARY

Cost Control

Set spending limits to prevent runaway costs:

max_request_spending_threshold = 0.50  # Max $0.50 per review

Or use a cheaper model for initial triage:

octomind run -m openrouter:openai/gpt-4o-mini "Quick check for obvious bugs"

Key Points

• --format plain or --format jsonl for non-interactive output
• Pipe input via stdin for non-interactive mode
• --model to balance cost vs quality per pipeline step
• Octomind has full tool access -- it can read files, not just the diff you pipe in

Note: Structured output with JSON schemas is available via the WebSocket and ACP interfaces. See Structured Output for details.