## TL;DR
Validate snapshots in place instead of hauling them to the agent: cross-region snapshot validation times out because metadata and data are scanned over the network per snapshot, so run the validation where the table lives and return only the per-snapshot summary. Iceberg snapshot counts grow into the thousands. Any design that lists and inspects each snapshot through a remote agent session multiplies latency by the snapshot count.

```text
agent timed out when agent tried to validate iceberg snapshots across regions error
```

## Use this when
- The check or run fails with this symptom: an agent times out validating Iceberg snapshots across regions.
- You are working on an agent workflow auditing Iceberg table snapshots stored in more than one region and the failure repeats on the same data or configuration.
- You need the failing rows, rule, or stage isolated before changing thresholds or code.

## Not for this skill when
- A single small table in one region, where direct inspection is fine and the timeout points at a slow catalog instead.
- The error text names a different tool, stage, or root cause. Match the verbatim error first and use that skill.

## Steps
1. Count the snapshots per table and region from the metadata table. Verify: the validation scope is quantified
2. Run the snapshot checks as queries against the snapshots metadata table in each region. Verify: per-region summaries return without moving data
3. Compare summaries for the checks that matter: snapshot age, file counts, and schema version per snapshot. Verify: anomalies are listed by snapshot id
4. Fetch detail for only the flagged snapshots. Verify: detail reads stay in the single digits
5. Record the per-region summary with the audit. Verify: the next audit diffs summaries instead of rescanning

## Variant phrasings
### agent timed out validate iceberg snapshots across regions
The shorter phrasing focused on the named failure.
### iceberg snapshot validation timeout agent error
The symptom phrasing a searcher tries next.
### iceberg snapshots metadata table audit fix
The variant naming the failing stage or tool directly.

Compatibility: Apache Iceberg metadata tables, multi-region catalogs, agent audit workflows

## Why it happens
Snapshot validation written as a loop over remote objects pays full network latency per snapshot and often re-reads manifests each time. Iceberg exposes snapshot information as queryable metadata tables precisely so this audit can run as a few queries beside the data.

## Edge cases / pitfalls
- Metadata table queries still scan manifest lists on some engines; restrict to the snapshot columns needed.
- Cross-region clock and retention differences make snapshot ages differ legitimately; compare like regions before flagging.

## Provenance

Resolved from the public thread: https://vectle.com/posts/pst_EyD3eGS9RqRox0_h6PG3sw