Learn

Everything you need to know to get started with RecursaAI on Solana

Quick Start

New to RecursaAI? Here is the fastest way to get started:

01

Connect your wallet

Click the Connect Wallet button in the top right corner. Supports Phantom, Solflare, and more.

02

Browse strategies

Explore yield strategies across JupLend, Kamino, Project0, Save Finance, Lulo, Loopscale, Exponent Finance, Meteora, Orca, Raydium, Jupiter Perps, Flash Trade, and Phoenix/Rise. Filter by risk, APY, or token.

03

Create a position

Select a strategy, set your leverage, and create a loop position with a single transaction.

Operator runbooks

Operator + developer-depth guides for every shipped feature. Each card opens the full Markdown on GitHub.

Operator runbooks

What you run when you're on-call. Health probes, deploy steps, mainnet gates, HA plan.

Monitoring & alerting

Service /healthz probes, alert routing, on-call runbook for keeper + nginx + frontend.

Ops deploy runbook

Hetzner sidecar + nginx + keeper systemd sync steps. Run this before every production deploy.

Dependency audit posture

Npm audit baseline policy for high / critical advisories across root, app, and keeper lockfiles.

Incident playbooks

On-call response paths for keeper-wallet drained, Pyth stale > 60s, and admin-key compromise. Symptoms, containment, recovery, post-mortem.

Mainnet pre-flight checklist

The list gating mainnet deployment. Every item links to the code / config to review.

Production environment inventory

Env vars, hot keys, sidecar URLs, registry IDs, oracle sources, smoke wallets, and emergency-control owners for mainnet.

App Cloudflare cutover

Issue #492 gate sequence for mainnet app env, server-side RPC proxy, Cloudflare KV, runtime smoke, and credential-leak evidence.

Keeper sidecar cutover

Issue #491 gate sequence for keeper, Pyth poster, Kamino farms, live health, wallet balances, restart, and rollback evidence.

Environment matrix

Per-network app, keeper, sidecar, and SDK env vars for switching between devnet, testnet, and mainnet without source edits.

Mainnet registry verification

Source-backed program ID, key account, on-chain executable, and close-verification evidence for every mainnet integration registry slot.

Oracle token LUT inventory

Mainnet pricing, token allowlist, Pyth feed, lookup-table, and inventory checker runbook for launch readiness.

Lending mainnet readiness

Launch-readiness ledger for JupLend, Kamino, Save, MarginFi Native, and Project0 account inventory, smoke gates, and execution blockers.

Lulo mainnet readiness

Issue #495 launch ledger for Lulo Protected and Boost smoke evidence, env gates, USDC-only mint policy, and $25k cap accounting.

Expanded surface readiness

Issue #458 repo-side closeout ledger for expanded strategy categories, child task status, release handoff issues, and production smoke gates.

Mainnet signoff evidence

Final launch evidence packet checklist and offline JSON checker for production deploy, smoke, config, and route signoff.

Mainnet deploy budget

Program deploy-wallet funding plan, rent calculation, balance separation, and evidence checker for issue #500.

Emergency controls

Mainnet pause, rollback, keeper rotation, registry disable, launch-cap, and communications rehearsal gate for issue #501.

Product video readiness

Colosseum recording readiness checklist for price freshness, explorer links, route claims, and demo-visible smoke gates.

RWA source verification

Mainnet source, pricing, eligibility, market-hours, and keeper-unwind decisions for every launch RWA and tokenized-equity route.

Colosseum demo script

Exact recording flow, route labels, fallback language, and evidence checklist for the Colosseum demo.

Security audit checklist

Internal pre-mainnet audit sign-off, launch-gating findings, and rules for filing follow-up security issues.

Mainnet day-0 smoke tests

The first transactions to run after the program is deployed. Tiny amounts, every instruction path.

Perps mainnet smoke

Issue #496 smoke runbook for Jupiter Perps, Flash Trade, and Phoenix/Rise entry, close, flat-state, and waiver evidence.

RWA demo smoke plan

Colosseum demo and mainnet smoke paths for Hastra, ONyc, xStocks, and gated treasury/fund RWA routes.

Tokenized equity QA runbook

xStocks and Ondo GM coverage map, approved-universe workflow, issue closure matrix, and mainnet smoke checklist.

Keeper + UI high-availability plan

Redundancy strategy for the hot paths so a single-node failure doesn't take positions offline.

Publishing to npm — runbook

Step-by-step for releasing @recursa-ai/idl, @recursa-ai/sdk, @recursa-ai/mcp. Covers one-time scope setup, version bumps, the release-npm.sh runner, post-publish verification, and rollback via npm deprecate.

Developer docs

How the thing is built. Test coverage map, adapter integration matrix, release-pass audit reports.

Testing coverage map

Which instruction / hook / component has which test. Grows with every feature.

Latest E2E audit report

Most recent end-to-end audit findings: gaps, fixes, deferred items.

E2E testing — runbook

How to run Playwright specs locally + against devnet. Default suite, devnet lifecycle opt-in, testid registry.

Adapter integration matrix

Which adapters (JupLend / Kamino / Project0 / Exponent / Jupiter Perps / Flash Trade / Jupiter Swap) are wired end-to-end vs stubbed.

Phoenix testing matrix

CI-safe Rise SDK builder fixtures, optional live Phoenix smoke tests, and the execution gate before enabling Phoenix-backed Recursa opens.

Post-Phoenix perp matrix

Venue-by-venue decision record for Flash Trade, GMTrade, Adrena, Pacifica, and Zeta/Bullet after Phoenix, with Drift explicitly excluded.

Native lending execution

Save Finance and MarginFi native execution classification, state-binding requirements, keeper gates, and follow-up implementation checklist.

RWA strategy research

Protocol and asset research for RWA yield support, pricing sources, risk boundaries, and demo-priority integrations.

Credix RWA evaluation

Decision record for Credix private-credit route feasibility, pass gating, CPI readiness, and when to revisit execution.

Known gaps tracker

Running list of gaps between code + docs + product. Mostly closes after every release pass.

Agent skill (Anthropic SKILLS.md)

Canonical capability doc for LLM agents (Claude Projects / MCP / LangGraph). Pairs with @recursa-ai/sdk + @recursa-ai/mcp.

Orchestrator architecture (phases 7-25)

Single reference for the open / close / keeper orchestrator: path matrix, LUT strategy (Recursa + Kamino + Jupiter), ComputeBudget calibration table, telemetry pipeline + dashboard, multi-tx fallback, ops runbook.

What is looping?

Looping is a DeFi strategy that multiplies your yield by creating a recursive borrow-and-supply cycle. You deposit collateral into a lending protocol, borrow against it, then deposit the borrowed funds as additional collateral. This cycle repeats to achieve leveraged yield exposure.

Step-by-step process

1. You deposit collateral (e.g., 1 SOL) into a lending protocol. 2. The protocol allows you to borrow against your deposit. 3. The borrowed assets are swapped back to your collateral token. 4. The new collateral is deposited as additional collateral. 5. Steps 2-4 repeat until your target leverage is reached. RecursaAI automates this entire process in a single transaction on Solana.

Understanding leverage

Leverage multiplies both your yield and your risk. A 2x loop means your effective position is twice your initial deposit. Higher leverage means higher potential APY but also a higher chance of liquidation if the market moves against you.

Health factor explained

The health factor measures how safe your position is. Above 2.0 is considered safe. Between 1.5 and 2.0 is moderate risk. Below 1.5 means your position is approaching liquidation. RecursaAI monitors your health factor, alerts when your guard is breached, and flags when manual deleverage is needed.