A manufacturer specification, retailer listing or policy statement tied to the exact model and source.
Every recommendation should show what is known, inferred and still missing.
Panlu is an editorial decision system, not a hands-on test lab. We combine traceable price observations, official specifications, timestamp-level independent measurements and explicit assumptions. High-impact calls stay unpublished or are downgraded when the evidence is incomplete.
A model-level result with value, unit, test condition and timestamp or directly locatable passage.
Panlu's interpretation of source facts or measurements. It must state the claim boundary and cannot masquerade as a measured result.
A declared input used for TCO or scenario math. Users should be able to see and, where supported, change it.
Requires exact-SKU source evidence, a saved evidence snapshot, checked and reviewed timestamps, a decision-specific reason, and either a comparison anchor or repeated price observations. A weak BUY is withheld from Deals and shown only as a reference anchor.
Requires a meaningful comparison anchor or repeated reviewed observations that show timing context. Ownership-only reference prices remain in Price Tracker.
Requires source-backed evidence of overpricing, price creep or a materially weaker same-purpose bundle. SKIP is not a negative product review.
One observation: a checked reference price, never a trend. Two to 29 reviewed observations in the 90-day window: thin_history; range and count may be shown, but confidence remains limited. Thirty or more reviewed checks: the operational 30-day spine is considered established.
Sampling records the retailer, exact SKU, price, availability, source URL, timestamp and evidence snapshot. Conflicting observations are retained when they represent different retailers, bundles or times; unexplained same-SKU conflicts go to review rather than being averaged away.
Affiliate economics are downstream of editorial decisions. Commission is excluded from verdict and offer ordering; transaction-ready offers require exact SKU, included items, seller authorization, fulfillment, returns, warranty and total checkout cost.
Automated jobs may create freshness reports and review queues. They only write public facts from explicitly approved sources; uncertain model matches, video measurements and warranty interpretations remain behind a human evidence gate.
Material corrections preserve the new decision and a dated explanation. Report a price, model or evidence problem through the correction entry on About.
TCO costs now use explicit SKU, cost-kind and scaling fields; material decision and method changes are preserved in a dated public recommendation log.
BUY calls now require a source URL, evidence snapshot, check and review timestamps, decision-specific rationale, and comparison or repeated price evidence. Incomplete BUY records move to Price Tracker as withheld references.
My Garage now distinguishes M12/M18, LXT/XGT, ONE+/40V and DeWalt 20V/FLEXVOLT. Ambiguous legacy brand data requires user confirmation.
Publishing checks now report current, under-24-hour, stale and future timestamps, and block missing product thumbnails from silently entering public SKU coverage.
Full recommendation change log (4)
The summary no longer asserts a fixed direction; typed SKU, kit, battery and charger costs drive the displayed comparison.
The starter kit was being inferred from its English label and scaled as tool-only cost. Typed cost categories removed the contradiction.Finder starts from none, Compare waits for explicit or Garage context, and Price Tracker starts with all references.
A default brand pair can be mistaken for an editorial recommendation before the user provides context.Expired or unverified action claims fall back to reference-only status until deterministic reverification succeeds.
Public BUY, WAIT and SKIP language must not outlive the evidence supporting it.Each now has a navigable hub with official source identity, an exact anchor and explicit performance or price gaps.
Users need a stable route into these lanes without Panlu pretending early coverage is a mature deal recommendation.