[Blockaid] Allow reporting false positives

[katspaugh]

What is the feature about

Users sometimes complain that Blockaid is giving them false positives for benign transactions.
Blockaid supports reporting such cases via the API after which Blockaid reviews them.

Allow the user to report false scans with a click of a button.

The list of requirements

• Add a button near the malicious tx warning saying "Report false result"
• Send a request to the COR-636
• ~~Send an API request to report the scan. Docs: ~~https://docs.blockaid.io/reference/evm-address-report
  • A workaround would be to prefill a UI developed by Blockaid, just like Metamask does it (see Linked Slack)

Additionally:

• When the user gets a malicious tx warning but clicks the checkbox "Proceed anyway", also report it to Blockaid via another request: https://docs.blockaid.io/reference/scan-status with the status accepted
• When the user abandons a tx with a warning, report it with the status rejected

Design

Open=Yes, Type=Threat, Content=High risk detected - Tx Assistance (Figma)

Transaction_Review - Transaction flow (Figma)

• Simon Lelievre (@simon) ask blockaid if we can pass parameters to the Blockaid form https://report.blockaid.io/mistake
  https://5afe.slack.com/archives/C05UN2C71RV/p1754935876173429

[katspaugh]

Tanya (@TanyaEfremova) could you prepare a design for this?

[the-real-chrizzo]

The reporting feature is not relevant right now as we are evaluating another solution. Happy to resurface this at a later point.

[mulligan-madness]

Heya wanted to check if you have any updates here... Had someone in our orgs address pop up as malicious even though litterally nothing has happened and they are still in possession of the multisig. Noticed in the gnosis ui. Even a link or something to know what actually caused the flag...

As it stands this is actually counterproductive for security because you are teaching people that "malicious" tags are inaccurate and can "usually" be safely ignored. Of course the onus is always on the user to verify, but presenting unqualified false information with no recourse to correct the error will only hurt.

[katspaugh]

cc the-real-chrizzo

[simabeh]

Hi Brennan Mulligan (@mulligan-madness),

We’re revamping how we display threat analysis, along with new features such as recipient verification and contract verification. Users will also be able to report false positives with a single click. We aim to have this live in Q4.

In the meantime, I recommend reporting the false positive directly through our support channel at https://help.safe.global/en/ (chat icon at the bottom right) or via the Blockaid portal https://report.blockaid.io/

The Blockaid team is usually quick to fix confirmed false positives, which should resolve your issue while we work on the new version.

Apologies for the inconvenience.