diff --git a/.claude/skills/audit-the-list/SKILL.md b/.claude/skills/audit-the-list/SKILL.md
index 57285d1d..961b3c7f 100644
--- a/.claude/skills/audit-the-list/SKILL.md
+++ b/.claude/skills/audit-the-list/SKILL.md
@@ -16,7 +16,7 @@ Resolve the scope from the arguments. Named sections mean exactly those, whether
 
 Fetch live evidence for every entry in scope before judging anything (CLAUDE.md verification rule):
 
-- **Downloads/month**: `cd website && UV_PYTHON=3.13 uv run python fetch_pypi_downloads_via_clickpy.py` — free keyless ClickPy sweep of the full README, sole writer of `data/pypi_downloads.tsv` (rewritten from scratch each run). Cross-checks print to stdout, take explicit names, and never touch the cache: `fetch_pypi_downloads_via_bigquery.py <name> ...` (canonical source, maintainer's own GCP account, `--dry-run` first — the docstring carries the cost constraints; full-README sweeps exceed the free tier, keep name lists small), `fetch_pypi_downloads_via_pepy.py <name> ...` (needs `PEPY_TECH_API_KEY` in repo-root `.env`, throttled to 5 requests/minute), or `https://pypistats.org/api/packages/{name}/recent` paced 8s or slower. pypistats excludes mirror/CI traffic; ClickPy, BigQuery, and pepy include it — never mix sources within one comparison.
+- **Downloads/month**: `cd website && UV_PYTHON=3.13 uv run python fetch_pypi_downloads_via_clickpy.py` — free keyless ClickPy sweep of the full README, sole writer of `data/pypi_downloads.tsv` (rewritten from scratch each run; header row, every row stamped with its `fetched_at` date). A cache whose `fetched_at` is within the last 7 days is current enough for verdicts — skip the sweep; older than that, re-run it (costs ~1s). Cross-checks print to stdout, take explicit names, and never touch the cache: `fetch_pypi_downloads_via_bigquery.py <name> ...` (canonical source, maintainer's own GCP account, `--dry-run` first — the docstring carries the cost constraints; full-README sweeps exceed the free tier, keep name lists small), `fetch_pypi_downloads_via_pepy.py <name> ...` (needs `PEPY_TECH_API_KEY` in repo-root `.env`, throttled to 5 requests/minute), or `https://pypistats.org/api/packages/{name}/recent` paced 8s or slower. pypistats excludes mirror/CI traffic; ClickPy, BigQuery, and pepy include it — never mix sources within one comparison.
 - **Repo state**: archived flag, last push, created date, stars, description — `gh api repos/{owner}/{repo}`, GitLab API for GitLab-hosted projects.
 - **PyPI metadata** (`https://pypi.org/pypi/{name}/json`) wherever a name might not be the canonical package — ownership collisions and wrong display names surface here.
 
diff --git a/website/fetch_pypi_downloads_via_clickpy.py b/website/fetch_pypi_downloads_via_clickpy.py
index e88815ac..e384b365 100644
--- a/website/fetch_pypi_downloads_via_clickpy.py
+++ b/website/fetch_pypi_downloads_via_clickpy.py
@@ -12,7 +12,10 @@ excludes mirrors by default.
 
 This script is the sole writer of data/pypi_downloads.tsv and rewrites it
 from scratch each run, so entries removed from README.md drop out
-naturally; names with no PyPI rows are written as NOT_FOUND. Cross-checks
+naturally; names with no PyPI rows are written as NOT_FOUND. The file
+starts with a header row (name, downloads, fetched_at) and every row
+carries the sweep date: a cache fetched within the last 7 days is current
+enough for audit verdicts, so only re-run when older. Cross-checks
 against other sources (fetch_pypi_downloads_via_bigquery.py,
 fetch_pypi_downloads_via_pepy.py) print to stdout and never touch the
 cache.
@@ -21,6 +24,7 @@ Usage: python fetch_pypi_downloads_via_clickpy.py
 """
 
 import re
+from datetime import date
 from pathlib import Path
 
 import httpx
@@ -67,7 +71,9 @@ def main() -> None:
     names = collect_names(README_PATH.read_text())
     print(f"Querying {len(names)} package names...")
     counts = fetch_clickpy(names)
-    OUT_FILE.write_text("\n".join(f"{name}\t{counts.get(name, 'NOT_FOUND')}" for name in names) + "\n")
+    fetched_at = date.today().isoformat()
+    rows = "\n".join(f"{name}\t{counts.get(name, 'NOT_FOUND')}\t{fetched_at}" for name in names)
+    OUT_FILE.write_text(f"name\tdownloads\tfetched_at\n{rows}\n")
     print(f"Done. {len(counts)}/{len(names)} names found on PyPI. Cached to {OUT_FILE}")