01 / 10

Solana Mobile Builder Grant · Developer Tooling

LaunchRig

LaunchRig is the release test rig for Solana Mobile apps.

A wallet flow can authenticate and still fail on the way back.

Working local prototype $28K milestone ask 12 weeks

Published evidence summary

PASS reference MWA 3 / 3

EXPECTED FAIL rejection broken

PASS rejection fixed

EXPECTED FAIL stale auth broken

PASS stale auth fixed

EXPECTED FAIL process death broken

PASS process death fixed

evidence: 9 reports / 18 checks

02 / 10

The origin and the problem

The failure happens after the wallet says done.

I wanted to build for Solana Mobile without a Seeker, on a laptop that could not comfortably carry an emulator-heavy workflow. Testing on an ordinary USB-connected Android phone exposed the larger problem: successful authentication did not prove safe recovery from every mobile lifecycle path.

R01

User rejectionThe app must recover without a false success state.

R02

Stale authorizationThe session must renew or fail clearly.

R03

Process deathRecover to UNKNOWN and enable retry.

03 / 10

One versioned release contract

One release gate for wallet recovery.

01launchrig.yml

Declare app identity, wallet fixture, network, core flows, and privacy policy.

02Validate

Reject unsafe or incomplete release configuration before device access.

03Doctor

Check the authorized Android device, APK bytes, wallet, and runner.

04Run suites

Use run for reference flows and matrix for controlled recovery pairs.

05Inspect evidence

Emit bounded HTML, JSON, and JUnit results. Public checksums verify file integrity.

Current config rejects mainnet Exact artifact checks Identifiers redacted Raw device evidence stays private

04 / 10

Working proof, not a concept

Broken recovery fails. The paired fix passes.

EXPECTED FAILRejection broken

PASSRejection fixed

EXPECTED FAILStale auth broken

PASSStale auth fixed

EXPECTED FAILProcess death broken

PASSProcess death fixed

3 × PASSconsecutive reference MWA suites

9normalized reports

18scenario checks

Two allowlisted builds were byte-identicalInternal verifier passed

One ordinary physical Android 12/API 31 profile. Self-recorded, unattested supporting technical evidence only.

Inspect the normalized proof Read the recovery case study

05 / 10

Why now

More mobile apps raise the QA burden.

100K+

power users exploring apps through the Solana dApp Store

Official grant guidance · 14 Jul 2026

1,000+

live apps in the Solana dApp Store

Official ecosystem update · 14 Jul 2026

Inference from official growth figures

As more apps ship to more mobile users, a shared Solana-specific release contract becomes more valuable than another one-off test checklist.

The current grant brief explicitly includes infrastructure and developer tooling.

06 / 10

Why Solana Mobile

MWA recovery today. Seeker next.

Established now

Ordinary Android + development fixtures

Reference and Mock MWA mobile wallet flows
Wallet rejection, stale authorization, and process-death recovery
Reference requests use testnet; controlled cases use devnet
Artifact identity, bounded logs, and offline verification

Proposed post-award

Actual Solana Mobile hardware qualification

Acquire or obtain separately permissioned Seeker access
Test production-wallet behavior as a distinct compatibility layer
Observe Seeker and Seed Vault-backed wallet behavior through MWA
Publish only separately permissioned, attributable results

MWA is the open app-to-wallet protocol on Android. Solana Mobile documentation directs dApp developers to MWA, while Seed Vault is primarily consumed through wallets. Sources: MWA and Seed Vault, checked 23 Sep 2026.

07 / 10

The missing layer in the stack

The missing release decision layer.

Layer	Generic Android automation	Mock MWA	Manual QA	LaunchRig
Execute Android UI flows	Core purpose	Not its purpose	Human-driven	Orchestrates a runner
Provide a development wallet fixture	Bring your own	Core purpose	Optional	Pins and integrates it
Version LaunchRig's mobile release policy	Project-defined	Wallet behavior only	Checklist-dependent	MWA, lifecycle, artifacts, scope
Emit privacy-limited, offline-checkable proof	Project-defined	Not its purpose	Review-dependent	One normalized format

This is a product-scope comparison, not a claim that every tool lacks extensibility. LaunchRig composes the execution and wallet layers into a Solana-specific release decision.

08 / 10

Founder-led execution

Shipped engineering is the execution record.

01

Working TypeScript CLIConfig validation, device doctor, run, matrix, reports, and proof verification.

02

25-rule local foundationStable rule definitions with executable positive and negative fixture coverage.

03

Public source checks301 CLI passes, 2 optional APK skips, and 56 fixture app passes.

04

Physical-phone evidence pipelineNine normalized reports and public checksums. Semantic verification remains internal.

Team boundary

Execution is currently founder-led and demonstrated through the working repository. The applicant chose to list no additional team member until a role and participation are confirmed. No unsupported biography or credentials are asserted.

The internal baseline recorded 319 LaunchRig tests and 56 fixture tests. The sanitized public checkout has 301 passing CLI tests and two optional cached-APK skips, plus 56 fixture passes. These counts describe different source scopes and do not imply new phone testing.

The separate Android Preview is an informational companion, not the CLI runner or the controlled fixture. It was installed and checked on one ordinary Android phone. Source is public, but no preview APK or dApp Store listing is published.

09 / 10

A 12-week public-good conversion

Each tranche closes on inspectable evidence.

M1Weeks 1-3

Publish the contract and qualify actual hardwareSupported 25-check release, public fixtures and Action, compatibility policy, Seeker access, separately labeled production-wallet and Seed Vault findings.

$8K

M2Weeks 4-6

Run permissioned publisher pilotsComplete three separately consented projects, publish operational measures, and report the 10-app and 100-run targets as achieved or missed.

$11K

M3 + M4Weeks 7-12

Public beta, review, v1, and handoffReport participation against the 25-project target, publish three approved case studies, complete independent review, and ship v1.0.0 with a maintenance guide.

$9K

Open ecosystem output: repository, rule contract, fixtures, validation Action, compatibility policy, evidence format, case studies, and maintenance guidance. All post-award figures are proposed milestones or targets.

10 / 10

The ask

$28K

A milestone grant to turn a working local release contract into a supported public tool validated on actual Solana Mobile hardware.

Fund the step from controlled ordinary-Android proof to open, permissioned, hardware-qualified ecosystem infrastructure.

Visit LaunchRig · Watch the demo · Inspect the source

Use of funds

Engineering$19,500

Devices and testing$3,000

Security and privacy review$2,500

Documentation and pilots$2,000

Contingency$1,000

Total · 12 weeks$28,000