Bootspring
HomeGet Started

Adaptive Workflows

Automatic failure detection and remediation paths for resilient workflows

Adaptive workflows automatically detect failures during execution and provide remediation paths to recover without abandoning progress.

How It Works#

When a workflow phase fails, Bootspring:

  1. Detects the failure type (testing, security, performance, deployment, database)
  2. Suggests a remediation path with specific steps
  3. Guides you through recovery with the right agents
  4. Resumes the workflow once the issue is resolved
┌──────────────────────────────────────────────────────┐ │ Normal Workflow │ │ Phase 1 ───▶ Phase 2 ───▶ Phase 3 ───▶ Complete │ └───────────────────┬──────────────────────────────────┘ │ Failure ▼ ┌──────────────────────────────────────────────────────┐ │ Remediation Path │ │ Diagnose ───▶ Fix ───▶ Verify ───▶ Resume │ └──────────────────────────────────────────────────────┘

Supported Failure Types#

Testing Failure#

Triggered when tests fail during a workflow phase.

1REMEDIATION_PATHS['testing-failure'] = { 2 steps: [ 3 { action: 'analyze_failures', agent: 'testing-expert' }, 4 { action: 'fix_issues', agent: 'code-review-expert' }, 5 { action: 'rerun_tests', agent: 'testing-expert' } 6 ] 7}

Security Failure#

Triggered when security vulnerabilities are detected.

1REMEDIATION_PATHS['security-failure'] = { 2 steps: [ 3 { action: 'security_audit', agent: 'security-expert' }, 4 { action: 'remediate', agent: 'security-expert' }, 5 { action: 'verify', agent: 'security-expert' } 6 ] 7}

Performance Failure#

Triggered when performance metrics don't meet requirements.

1REMEDIATION_PATHS['performance-failure'] = { 2 steps: [ 3 { action: 'profile', agent: 'performance-expert' }, 4 { action: 'optimize', agent: 'performance-expert' }, 5 { action: 'benchmark', agent: 'performance-expert' } 6 ] 7}

Deployment Failure#

Triggered when deployment fails.

1REMEDIATION_PATHS['deployment-failure'] = { 2 steps: [ 3 { action: 'diagnose', agent: 'devops-expert' }, 4 { action: 'fix_config', agent: 'devops-expert' }, 5 { action: 'retry_deploy', agent: 'devops-expert' } 6 ] 7}

Database Failure#

Triggered when database operations fail.

1REMEDIATION_PATHS['database-failure'] = { 2 steps: [ 3 { action: 'analyze_schema', agent: 'database-expert' }, 4 { action: 'fix_migration', agent: 'database-expert' }, 5 { action: 'verify_data', agent: 'database-expert' } 6 ] 7}

Using Adaptive Workflows#

Reporting a Failure#

# Report a test failure bootspring workflow report-failure testing # Report with details bootspring workflow report-failure security --details "SQL injection in user input"

Starting Remediation#

1# Start remediation for a failure 2bootspring workflow remediate <failure-id> 3 4# Output: 5# Remediation started. First step: Analyze test failures 6# Agent: testing-expert

Advancing Through Remediation#

1# Complete current step and move to next 2bootspring workflow remediation-next 3 4# Output: 5# Step 2/3: Fix failing code 6# Agent: code-review-expert

Completing Remediation#

1# After all steps 2bootspring workflow remediation-next 3 4# Output: 5# Remediation complete. You can resume the workflow. 6bootspring workflow resume

API Usage#

1const orchestrator = require('bootspring/intelligence/orchestrator'); 2 3// Report a failure 4const failure = orchestrator.reportPhaseFailure('testing', { 5 message: 'Integration tests failing', 6 errorCount: 3 7}); 8// { success: true, failureId: 'failure-123', remediation: {...} } 9 10// Start remediation 11const remediation = orchestrator.startRemediation(failure.failureId); 12// { success: true, currentStep: { action: 'analyze_failures', agent: '...' } } 13 14// Advance through steps 15orchestrator.advanceRemediation(); 16// { success: true, complete: false, stepNumber: 2 } 17 18// Complete remediation 19orchestrator.advanceRemediation(); 20// { success: true, complete: true, message: 'Remediation complete' } 21 22// Resume workflow 23orchestrator.resumeWorkflow();

Failure History#

View past failures and their resolutions:

const failures = orchestrator.getWorkflowFailures(); // [ // { id: 'failure-123', type: 'testing', resolved: true, resolvedAt: '...' }, // { id: 'failure-124', type: 'security', resolved: false } // ]

Best Practices#

  1. Be specific - Include details when reporting failures
  2. Follow the path - Trust the remediation steps in order
  3. Verify before resuming - Ensure the fix actually works
  4. Learn from failures - Check failure history for patterns

Related#