Repository intelligence · Live beta

Know what you’re shipping.
Before release.

Review JavaScript and TypeScript projects built or accelerated with AI before release. Trace production flows, observed security controls, and engineering risks from repository evidence without executing source code or asking AI to decide the findings.

FlowPreflight analysis screen showing pipeline progress at 8 of 37 steps, service and queue status, estimated time, and evidence for each step
REAL ANALYSIS IN PROGRESS37 STEPS · LIVE EVIDENCE · NORMAL / BOOST
ZIP / RAR / TARPUBLIC GITHUB REPOSITORY4 PUBLISHED REAL RESULTS24-HOUR RESULT RETENTION

01 / PRODUCT

Review the system,
not a sample of files.

Start from an archive or public GitHub repository and inspect four published results from real repositories before submitting your own source.

01 / INPUT

Start from the source you choose

Upload a ZIP, RAR, or TAR up to 25 MiB or submit a public JavaScript/TypeScript GitHub repository, then choose Normal or Boost mode.

02 / ANALYZE

Review the complete supported scope

Deterministic analysis links files, routes, symbols, function calls, controls, data operations, and provider boundaries without executing source code.

03 / DELIVER

Receive evidence and executable work

Get a Customer Report, Technical Flow Dossier, and Prompt Pack that keep facts, recommendations, unknowns, and analysis limitations distinct.

02 / ANALYSIS METHOD

Repeatable evidence,
not an AI opinion.

The core engine examines every file in the supported static scope, preserves every detected flow, and keeps limitations and unresolved boundaries explicit in the artifacts.

Every output traces back to source and graph evidence. Facts, recommendations, analyzer gaps, and unknowns remain distinct before the artifacts pass reconciliation and quality checks.

01

Inventory the supported scope

Enumerate supported JavaScript, TypeScript, manifests, and configuration while keeping exclusions, limits, and parse failures explicit.

02

Resolve how code connects

Link imports, symbols, calls, routes, inputs, controls, data operations, provider calls, and side effects where static resolution succeeds.

03

Reconstruct every detected flow

Trace entrypoint-to-result paths. When a path cannot be closed, retain the evidence and exact analysis boundary instead of guessing.

04

Apply versioned knowledge

Compare flows with engineering checklists for identity, payments, webhooks, data writes, queues, AI providers, and other detected capabilities.

05

Validate every published claim

Reconcile counts and references, distinguish facts from recommendations and unknowns, then run privacy, contradiction, and artifact-quality gates.

06

Keep AI outside the decisions

AI may translate or improve wording after review, but cannot add findings, change severity, hide unknowns, or alter Prompt Pack tasks.

FROM SOURCE TO VERIFIED OUTPUTNO CODE EXECUTION · TRACEABLE · EXPLICIT BOUNDARIES
Repository inputArchive up to 25 MiB or public GitHub URL · Normal / Boost
Deterministic reviewInventory · resolve · reconstruct flows · apply versioned knowledge
3 verified artifactsCustomer Report · Technical Flow Dossier · Prompt Pack

LIVE BETA AVAILABLE NOW

Release with context,
not assumptions.

FlowPreflight is ready for real-world testing. Analyze a repository, review evidence-backed outputs, and help shape the product with your feedback.

Start testing ↗