For Salesforce
Your org is metadata in folders. AI needs it as features.
IngestMD groups LWC, Apex, and objects into the features they form, following @salesforce/apex calls and SOQL into your data model. The result is AI-ready context for reviews, refactors, onboarding, and handover, mapped on your laptop in about a second.
The problem
Salesforce metadata fights AI tools
A Salesforce project is hundreds of small files spread by metadata type: a component here, its Apex there, the object somewhere else. Flatten the whole folder into an AI tool and the model gets a flat list with no sense of which pieces form a feature. And client orgs hold data and IP you cannot paste into a cloud chat in the first place.
What IngestMD does
It reads your org the way you think about it
Groups by feature, not folder
Components, custom objects, Apex, Lightning pages, and supporting metadata clustered into the features they form.
Follows the wiring
Tracks @salesforce/apex calls and SOQL from a component into the objects it touches, so an area is a real feature.
Deterministic and offline
The same map on every run, fully on your machine. No org connection required, nothing uploaded.
Reads Apex with SOQL
Apex, LWC, Aura, and objects, parsed and indexed alongside the rest of your stack.
Real runs
Public Salesforce projects, mapped on a laptop
LWC Recipes
Apex / LWC- Lightning Web Components393
- Custom Objects31
- Lightning Pages25
- Other Metadata25
Apex Recipes
Apex / LWC- Other Metadata86
- Apex Classes74
- Custom Objects45
- Lightning Web Components23
DreamHouse
Apex / LWC- Lightning Web Components51
- Custom Objects38
- Other Metadata19
- Apex Classes9
E-Bikes
Apex / LWC- Lightning Web Components49
- Custom Objects40
- Other Metadata27
- Apex Classes9
What you do with it
Built for the work, not just the bundle
AI code review of Apex and LWC
Bundle a feature, its tests, and the objects it touches, then hand it to AI for a review that sees the whole picture instead of one class.
Plan a refactor
Discover the area that owns a flow, see its components, Apex, and data model in one place, and reason about the blast radius before you touch it.
Onboard onto an org
Hand a new admin or developer a named map of the org, grouped by feature, so their AI assistant answers questions grounded in the real metadata.
Prep a handover or audit
Turn a client org into a clean, redacted set of Markdown a reviewer or a model can actually read, with secrets scrubbed.
For client work
Local by default, so consultancy code stays put
Discovery, bundling, and redaction run on your machine, offline, with no account and no telemetry. Secrets are scrubbed before anything is written. The only network egress is opt-in: your own local model, an optional cloud model with your key in the OS keychain, or Confluence retrieval. A client org can finally get the AI treatment without leaving your laptop.
Map your org in about a second
Free to download, with a 3-day full Pro trial. No account, no card, nothing uploaded.