Active shards in numeric order, then closed archive in reverse path order when enabled; original row order; stop at target. Convenience sample, not representative.

Analysis snapshot: Includes active and closed archived postings

Immutable source revision: 7675a78a63ea3d42d47515a350ac7650da972bd0

Source snapshot: 10/25/2026, 4:29:33 PM UTC. Dates in the chart are provider-supplied posting dates, not collection dates.

• Verified aggregate snapshot, not a live job feed.
• Convenience sample; monthly source dates do not measure market-wide hiring growth.
• Country-only location; no city, salary or role-family breakdown is inferred.
• Skills overlap across postings; skill counts cannot be summed to count distinct jobs.
• Alias matching can be ambiguous (for example, go); skill precision is not audited.
• Duplicate candidates are heuristic decisions, not audited true duplicates.
• Latency is local, sequential and cached; it is not a production SLA.

Skill percentages use all retained postings in the selected months as the denominator. Category and text filters narrow the displayed skills; they do not change that denominator. Skill mentions overlap. Search and category filters do not change the monthly posting chart.

API timing is measured locally with 1 sequential client. It is not the speed of this dashboard or a production service guarantee.