Taskeen Haider
Founder / GitHub Organization Owner
GitHub profile — Taskeen Haiderraitaskeen@legacyexodus.devVerification-driven software modernization · Active engineering foundation
Legacy systems shouldn't
dictate your future. Trace their behavior.
Modernization begins with understanding. LegacyExodus is building a governed migration platform that maps complex codebases, establishes traceable engineering evidence, and lays the foundation for controlled transformation and independent verification.
JavaScript & TypeScript analysis available in the engineering foundation.
Automated migration and verification are on the roadmap.
Automated migration and verification are on the roadmap.
DETERMINISTIC BY DESIGN JAVASCRIPT / TYPESCRIPT SOURCE-TRACEABLE EVIDENCE
The modernization problem
Why enterprise migrations fail.
And why understanding comes first.
Modifying mission-critical software without structural ground truth is the primary driver of costly rewrite regressions. Systems hold decades of essential logic in undocumented behaviors that traditional linters miss and unconstrained AI hallucinates.
Undocumented runtime logic
Decades of accumulated edge cases live in runtime assumptions, unwritten side effects, and tacit team conventions rather than specifications. Greenfield rewrites discard this operational knowledge.
Sprawling dependency coupling
Deeply nested imports, shared mutable state, dynamic dispatch, and framework lifecycle hooks cross module boundaries in ways that surface only during production outages.
Unbounded AI hallucination
Unconstrained LLMs invent non-existent APIs, drop critical validation branches, and produce plausible code that fails silently in production without verifiable ground truth.
The LegacyExodus methodology
Discover → Understand → Transform → Verify
LegacyExodus replaces unguided rewrites with a governed four-stage modernization pipeline. The first two stages provide an active, deterministic engineering foundation; the latter two govern automated transformation and independent verification.
FOUR-STAGE GOVERNED MIGRATION ARCHITECTURE
Verified Foundation Partial Foundation Target Architecture
STAGE 01 · DISCOVER THE SYSTEM
VyrixScout
IMPLEMENTATION STATUSVERIFIED
Inventory the repository without assumptions. Classify source files, identify workspace boundaries, resolve module topologies, and construct deterministic dependency maps before any code transformation begins.
Inputs
- Git repository or filesystem tree
- Package manifests (package.json, lockfiles)
- Configuration files (tsconfig.json, module resolvers)
Capabilities
- Deterministic repository inventory
- Source classification & file typing
- Module topology & workspace mapping
- Dependency signal extraction
- Package & monorepo resolution
Outputs
- Normalized source inventory and file registry
- Module topology and dependency adjacency graph
- Project structure classification and package boundaries
BUSINESS VALUE
Eliminates blind spots. Engineering leaders gain immediate, factual visibility into codebase scale, module coupling, and migration complexity before committing budget.
STAGE 02 · RECOVER THE ARCHITECTURE
NolaraStruct
IMPLEMENTATION STATUSPARTIAL
Extract ground-truth structural intelligence from code. Parse ASTs, model lexical scopes and symbols, reconstruct interprocedural call graphs, build control-flow and data-flow graphs, and compute conservative taint facts.
Inputs
- Source files validated by VyrixScout
- Normalized AST structures (Babel-parsed JS/TS)
- Resolved symbol import/export tables
Capabilities
- Babel AST parsing & normalization
- Lexical scope & symbol resolution
- Semantic call-graph reconstruction
- Control-Flow Graph (CFG) reachability
- Data-Flow Graph (DFG) value tracking
- Conservative taint-propagation facts
Outputs
- Lexical scope trees and symbol definition/use chains
- Control-Flow Graphs (CFG) and Data-Flow Graphs (DFG)
- Interprocedural semantic call graphs & conservative taint facts
- Structured JSON artifacts with SQLite / memory storage
BUSINESS VALUE
Recovers lost business logic and data propagation paths without runtime execution risk. Developers understand exact dependencies and branch reachability before touching legacy code.
STAGE 03 · ENGINEER THE TRANSFORMATION
ZelvoxForge
IMPLEMENTATION STATUSPLANNED
Transform verified evidence into modern code through bounded AI reasoning. The AI gateway receives immutable, source-grounded graph snapshots to generate candidate architectures and typed target code (Rust/Axum/SQLx and modern TypeScript).
Inputs
- NolaraStruct verified evidence graphs & taint boundaries
- Target architecture constraints & framework schemas
- Transformation contracts and semantic invariants
Capabilities
- Bounded AI reasoning with versioned graph context
- Structured transformation planning & contracts
- Target code generation (Rust/Axum/SQLx, modern TS)
- Semantic invariant preservation mapping
- Strict separation of AI proposals from acceptance
Outputs
- Candidate modernized target source code
- Traceable transformation plans with source citations
- Semantic diffs linking generated code to original AST nodes
BUSINESS VALUE
Replaces unguided, hallucination-prone AI code generation with disciplined, evidence-bounded transformation. Every proposed line references verified source semantics.
STAGE 04 · VERIFY BEFORE ACCEPTANCE
VymosGate
IMPLEMENTATION STATUSPLANNED
Independent verification and human governance. Evaluate candidate migrations in isolated compiler sandboxes, run differential tests against original behavior, enforce fail-closed verification gates, and require human architectural sign-off.
Inputs
- ZelvoxForge candidate generated code
- Original system behavior profiles & test suites
- Transformation verification contracts & compiler configs
Capabilities
- Isolated compiler sandboxing & build checking
- Differential testing against baseline test suites
- Behavioral equivalence & invariant verification
- Fail-closed acceptance gates
- Auditable human-in-the-loop governance sign-off
Outputs
- Sandboxed compilation & build verdicts
- Differential execution test reports & coverage diffs
- Verification audit trail for human release sign-off
BUSINESS VALUE
Zero unverified code enters production. Independent validation gates ensure the transformed system compiles, passes differential tests, and receives human authorization before deployment.
Deterministic analysis foundation
Reconstruct ground truth
before transforming code.
Inspect repository structure, call hierarchies, control-flow paths, and architectural boundaries directly from source. The active JS/TS engine recovers verifiable facts so modernization decisions are never based on guesswork.
Source repositoryJavaScript and TypeScript are the current analysis focus. This repository layout is schematic.
Repository intelligence
VyrixScout inventories files and reconstructs modules and dependencies. Discovery establishes the system boundary before deeper analysis.
inventory.json · modules · dependency relationships
classifyScoreRepresentative function name from the public sample; this AST is an explanatory schematic.
Structural understanding
NolaraStruct models syntax, scopes, and symbols. Stable structure makes program relationships inspectable instead of leaving them buried in text.
AST · lexical scopes · symbol relationships
Function entryEntry into classifyScore. This is static control flow, not an execution trace.
Control & data flow
Control flow exposes possible paths. Data flow concerns how values relate across program operations. The public sample demonstrates selected intraprocedural CFG and reachability; DFG belongs to the broader documented foundation.
DFG CONCEPT · BROADER FOUNDATION
scoreparameter>= 70comparison- Branch condition
Illustrative value dependency, not captured DFG output. The public CFG sample excludes data-flow analysis.
CFG schematic · both return branches structurally reachable · DFG excluded from public sample
Source revisionArtifacts need source provenance to remain interpretable.
Deterministic evidence
Canonical artifacts retain relationships and source provenance. Reproducibility lets engineers compare and revisit a model; it does not guarantee a complete model or behavioral equivalence.
canonical JSON · stable identifiers · source provenance
Engineering evidence
Grounded in live telemetry.
Zero vanity metrics.
We hold our platform to compiler-grade validation standards. Our architectural claims are backed by automated test suites, formal benchmark governance audits, and stratified ground-truth baselines.
1,537
Tests Passed
145 test files, 16,413 assertions, 0 failed, 1 skipped. Bun 1.4.2 runtime.
10 / 10
Benchmark Governance
10/10 automated checks PASS. SHA 97ab7a9dcddaf4a7 audit verdict.
25 Files
Sample Evaluation Corpus
Representative ground-truth fixtures from Express v4.21.2 & Fastify v5.3.3. Sample-based evaluation; not full framework coverage.
92 Tests
Public Engineering Sample
Five-file inspection slice with AST and CFG in legacyexodus-analysis-sample.
PUBLIC ENGINEERING SAMPLEReal source.
Real source.
Visible boundaries.
AST-driven control flow, reachability, graph invariants, and canonical serialization. A focused engineering slice you can inspect on GitHub.
function classifyScore(score: number) {
if (score >= 70) {
return "pass";
}
return "review";
}
classifyScore(82);Both return paths are structurally reachable. This public sample excludes DFG and interprocedural analysis.
Recover the bigger picture
Complexity accumulates.
Understanding shouldn't disappear.
Files multiply. Dependencies shift. Knowledge gets scattered. Before changing a system, recover the relationships that make it work.
Source files Dependencies Program structure Possible paths
Structured understanding.
Relationships, provenance, and canonical artifacts. A system model you can inspect.
SOURCE TO EVIDENCE / SCHEMATICData protection & control
Your source code stays in your perimeter.
LegacyExodus is designed with a strict zero-leakage, local-first security philosophy. We never require you to upload proprietary code to an unverified third-party cloud.
LOCAL-FIRST EXECUTION
Local analysis runtime
The analysis CLI operates entirely on your machines or internal CI runners. AST parsing, graph extraction, and artifact serialization remain within your security boundary.
FULL IP OWNERSHIP
Customer owns all artifacts
You retain exclusive ownership of your source repositories, generated structural graphs, and target migration candidates. No proprietary code trains shared public AI models.
ZERO CLOUD INGESTION
No marketing code intake
This website accepts no repository uploads, runs no analytics trackers, and operates as a static informational deployment. Discovery conversations occur via scheduled technical briefing.
The people behind the foundation
Built with conviction.
Founded August 5, 2026. Based in Pakistan. Three founders working toward software understanding and modernization that engineers can trust.
Shahzaib Hassan
Co-Founder & AI Engineer
GitHub profile — Shahzaib Hassanmehrshahzaib@legacyexodus.devAbbas
Co-Founder & Software Engineer
GitHub profile — Abbasbilobhatti@legacyexodus.devStart with understanding.
Follow the engineering, explore the approach, or get in touch.