The OrbitFlare Infra Accelerator is a 4-week intensive program. Each week has a clear focus and expected deliverable.
The first week focuses on understanding where each project stands today and defining a realistic path to mainnet.
- Welcome and program expectations
- Setup of a private communication channel between each team and OrbitFlare
- Establish check-in schedule (Tuesdays and Fridays)
Each team receives a dedicated session reviewing their codebase. We assess:
- Current state of on-chain programs
- Frontend and backend infrastructure
- External dependencies (oracles, APIs, indexers)
- Known technical debt from hackathon or prototype development
Based on the architecture review, we produce a mainnet blockers list. Common blockers include:
- Missing error handling or edge cases
- Hardcoded devnet addresses
- No transaction retry logic
- Missing input validation
- Lack of monitoring or logging
- Unchecked arithmetic in financial logic
- Missing access controls or authority checks
Each team defines 3-4 clear milestones for the remaining three weeks. Milestones must be specific and verifiable.
- Not: "Improve security"
- Instead: "Add checked math to token calculations and write overflow edge-case tests"
Week 1 Deliverable: Mainnet blockers list and milestone plan.
This is the primary development sprint. Teams focus on resolving mainnet blockers and integrating OrbitFlare infrastructure.
- Fix all items on the blockers list
- Add comprehensive error handling
- Ensure arithmetic safety checks
- Validate account constraints and authority checks
- Expand test coverage for adversarial scenarios
- Replace public RPC endpoints with OrbitFlare
- Implement proper transaction sending patterns
- Configure confirmation polling and retry logic
- Implement WebSocket subscriptions where required
- Configure gRPC streaming connections for Jetstream users
Teams configure production-grade infrastructure including:
- Staging environment mirroring mainnet configuration
- Monitoring of transaction success rates and RPC latency
- Logging sufficient to debug production issues
Week 2 Deliverable: OrbitFlare RPC integrated, mainnet blockers resolved.
Development continues but the focus shifts toward resilience and stability. If something can break in production, it should break during this week.
Teams simulate real-world load conditions:
- Burst transaction scenarios
- Concurrent users
- RPC timeouts and network congestion
- Slot skipping scenarios
- Duplicate transaction submissions
- Invalid inputs and expired blockhashes
We review projects for common Solana vulnerabilities including:
- Missing signer checks
- PDA seed collisions
- Reinitialization attacks
- Unsafe CPI calls
- Privilege escalation paths
Week 3 Deliverable: Stress testing complete, security review issues addressed.
Everything built in the previous three weeks culminates in a live mainnet deployment.
Before deploying:
- All blockers resolved
- Full test suite passes
- Stress testing completed
- Security review issues addressed
- Monitoring and alerting configured
- Program authority secured
- User documentation exists
- RPC endpoints pointed to OrbitFlare
OrbitFlare engineers are available in real time for:
- Program deployment assistance
- Test transactions with real SOL
- Verifying addresses and program IDs
- Monitoring the first hours of production traffic
- Flip the product live for users
- Monitor transaction success rates and logs
- Address issues that surface with real traffic
- Deploy hotfixes if required
Each project should publish a mainnet launch thread or announcement describing:
- The product and what it does
- The journey from prototype to mainnet
- Key challenges solved during development
- The infrastructure used to support the deployment
Each team receives a wrap-up call covering:
- What went well
- What could improve
- Remaining roadmap items
- Continued RPC access
- Introductions to relevant ecosystem partners where appropriate
Week 4 Deliverable: Live mainnet launch, public announcement published.
- Primary channel: Private Telegram or Discord channel per team
- Check-ins: Tuesdays and Fridays, 30 minutes each
- Async support: Available at all times
- Escalation: Direct access for critical blockers
- Cross-team: Optional interaction between cohort teams
The program is successful if by the end of Week 4:
- Each team has a working product deployed on Solana mainnet
- Each team is using OrbitFlare RPC in production
- Each team has monitoring in place and can respond to production issues independently
- Each team has a clear roadmap beyond the accelerator
- Each team publicly shares their launch and acknowledges the infrastructure powering their deployment