Update is now live. Applicable for new scans and coins.
Faster qualification: new coins get another market lookup when the first response is empty.
Better priorities: coins above $5k volume and $5k market cap get website discovery sooner.
Less queue starvation: unprocessed launches keep their place even as they get older.
Balanced polling: new launches, retries, known projects and older coins all retain capacity.
Core-focused scores: working support chat or optional statistics cannot stand in for the advertised product.
Clearer missing coverage: untested main claims stay visible instead of disappearing from the assessment.
Broader product testing: added a reusable test for browser-based market watchers and alerts.
Real input variation: the test changes thresholds and checks that the result changes correctly.
Negative controls: empty input, hardcoded answers, always-positive results and invisible outputs cannot earn a passing score.
Test cleanup: temporary watchers must be removed before that workflow earns credit.
More reliable form entry: a slow action can recover when the exact requested value was actually entered.
Better rendering waits: tests briefly wait for delayed data and readable outputs.
Reduced-motion testing: avoids decorative animation and smooth scrolling interfering with functional checks.
Better app navigation: follows safe, observed app and history links, including hash routes.
Better public-data access: supports narrowly approved public statistics requests without allowing database writes.
Automatic browser recovery: queued research can recover after startup failures, with controlled backoff.
Automatic historical checks: matching published routing receipts can be checked against confirmed transactions and official bridge records.
Stricter receipt attribution: an unrelated token’s receipt cannot automatically earn another project a score.
Cleaner cancellation: interrupted receipt requests release their connections.
More specific reviews: findings describe the attempted workflow, observed result and remaining limits.
Clear separation of evidence: peripheral features, historical results and manual score adjustments remain distinct.
Better diagnostics: interrupted tests retain bounded details that help identify the actual failure.
More regression coverage: added tests for scoring errors, queue delays, fake outputs, wallet activity and worker recovery.
Soon:
Faster market retries when a new coin’s first lookup returns no data.
Priority website discovery for coins meeting the $5k volume and $5k market-cap requirements.
Queue fixes so older, unprocessed launches don’t get stranded.
Preserved polling capacity for new launches, known projects and older tokens.
Scores based on demonstrated core functionality;
support chat and optional statistics stay separate.
Unsupported main claims remain explicitly unresolved.
A new market-watcher test that checks real inputs, outputs, invalid input and cleanup.
Threshold variation to catch static or hardcoded results.
Rejection of invisible outputs and wallet activity during read-only tests.
Bounded waiting for delayed app data.
Safer navigation through observed app and history routes, including hash routes.
Access to narrowly approved public statistics requests.
Automatic browser recovery with backoff when research is waiting.
Automatic historical receipt checks with strict identity matching, timestamps and read limits.
Shorter, specific reviews describing the attempted workflow and observed result.
New regressions for queue starvation, false scores, fake outputs and interrupted requests.
WatchLarp update 👁️
• Scores now focus on demonstrated core product functionality.
• Improved public-data access during product checks.
• Fixed generated test-wallet detection and selection.
• Added recovery for temporary browser-worker upload failures.
• Recent verified evidence carries across rescans.
• Clearer scan statuses, findings and test limitations.
• Audited admin score controls from 1–100, separate from automated evidence.
• Faster Pulse loading with shared snapshots and fewer repeated requests.
• Improved PFP and metadata recovery, with website/X links.
• Volume and market cap displayed as observations arrive.
• New pairs newest first; Qualified highest score first; Migrated newest graduation first.
• Research requests by https://t.co/5knsYHVXXw address from Pulse and Checks.
• Compact reviews, expandable evidence and downloadable reports.
The point is not to replace Axiom's pulse but to be an additional layer of it.
The reality is there are not that many non-larp projects.
It's the kind of thing you keep open on your second monitor all the time and check time to time.
While updating the website for the first fix, something changed in the score system that we had initially. That has been fixed. Update is now live.
However another more nuanced improvement is coming soon.
Update is now live.
Faster first checks for new tokens’ metadata and websites.
Better PFP loading and recovery from previously missing images.
Manual PF address research on both Pulse and Checks.
Reuse of recent checks to avoid duplicate work and costs.
Fair scheduling between manual requests and automatic scans.
Newer manual reviews appear alongside automatic evidence.
Public “Investigating” labels changed to “Checking.”
Show all coins remains off by default.
Pulse stops requesting data from other pages or hidden tabs.
Repeated updates share one active request.
New pairs sort newest first; Qualified sorts highest score first; migrations use graduation order.
Reviews focus on product evidence, blocked attempts and untested features.
Temporary browser failures use bounded retries.
Review details stop reloading on every price update.
Improved navigation for testing products such as Compute.
Keep in mind our pulse is not meant to replace Trading Terminal's scanning pages.
You can't buy or sell tokens (at least for now), the only way to use is to see if a coin is a a larp or not
We caught $Seed before anyone.
If a coin does not meet the requirements, you wont see unless you click show all coins.
Quality over quantity.
Up next:
Faster first checks for new tokens’ metadata and websites.
Better PFP loading and recovery from previously missing images.
Manual https://t.co/5knsYHVXXw address research on both Pulse and Checks.
Reuse of recent checks to avoid duplicate work and costs.
Fair scheduling between manual requests and automatic scans.
Newer manual reviews appear alongside automatic evidence.
Public “Investigating” labels changed to “Checking.”
Show all coins remains off by default.
Pulse stops requesting data from other pages or hidden tabs.
Repeated updates share one active request.
New pairs sort newest first; Qualified sorts highest score first; migrations use graduation order.
Reviews focus on product evidence, blocked attempts and untested features.
Temporary browser failures use bounded retries.
Review details stop reloading on every price update.
Improved navigation for testing products such as Compute.