A redacted proof packet for sandbox email risk
This example uses demo metadata only. It shows the object Silobase gives a Salesforce admin or consultant: path counts, recipient class, unresolved coverage, and the safe-test actions to finish before UAT email is opened.
| Source | Sender | Recipient class | Risk | Release note |
|---|---|---|---|---|
| Flow | Workflow context | Real-contact risk | High | Would-send path needs containment before UAT deliverability opens. |
| Apex | SingleEmailMessage | External literal | High | Literal address stays redacted; owner must confirm it is test-safe. |
| Approval | Approval alert | Unparsed endpoint | Review | Coverage gap is visible instead of hidden behind a green result. |
| Org-wide sender | Sandbox sender | Sender config | Review | Sender identity is checked before anyone treats the path as safe. |
Show the risk shape without showing private metadata
The proof packet keeps the row-level object a release owner needs, but replaces raw addresses and org details with recipient classes. A consultant can still point to the risk mix, source type, sender class, and unresolved endpoints.
- 8 email paths from demo metadata
- 2 free-preview findings and 6 gated report findings
- No raw email addresses, org IDs, tokens, or customer metadata
Risk mix
Demo scan / redacted counts
Unparsed paths stay visible
The demo scan does not pretend every branch is solved. Approval-process endpoints, literal Apex addresses, and sender configuration gaps stay on the page with a manual-review note, so the release team can decide before deliverability changes.
- Apex is treated as heuristic static review, not full symbolic execution
- Manual-review rows block clean sign-off until someone owns them
- Resolved and unresolved paths can both appear in the same proof packet
Source coverage
Detected paths by metadata family
Three paths require manual review before anyone signs off the sandbox for UAT.
Every finding maps to a next action
The paid report should not be a wall of warnings. It turns the manifest into a contained test plan: check the sender, inspect the recipient class, retest after the fix, and keep the result on the release ledger.
- Review approval alerts whose recipient endpoint could not be parsed.
- Inspect Apex SingleEmailMessage paths for literal addresses before UAT.
- Confirm org-wide senders point to intentional sandbox-safe identities.
- Retest after metadata changes and save the result as release evidence.
Safe-test actions
What the release owner does next
Review approval alerts whose recipient endpoint could not be parsed.
Inspect Apex SingleEmailMessage paths for literal addresses before UAT.
Confirm org-wide senders point to intentional sandbox-safe identities.
Retest after metadata changes and save the result as release evidence.
What this demo proof packet contains
- 8
- email paths mapped from fixture metadata
- 3
- unparsed endpoints left for manual review
- 0
- raw addresses or org identifiers exposed here
Proof boundary
What this page does not claim
This is not an official Salesforce artifact. It is not a guarantee that no email can send. It is a redacted example of how a metadata scan becomes release evidence.
Not affiliated with Salesforce
No guaranteed prevention claim
No raw customer metadata
No full Apex symbolic execution claim
Turn your own sandbox export into a proof packet.
Upload metadata, inspect the first findings free, then decide whether the $99 report or $39/mo ledger belongs in the release process.