emrg: reader_fix_latency.py — print n (sample size) + open-issue age counter

[how2how2how2-arch]

Summary

Implements issue #1056 (Dev.to comment 3djoa, heinrichneb, on the 50-minute-loop article): two improvements to scripts/reader_fix_latency.py.

Changes

1. Print n (sample size): the median line now reports n={len(samples)}, so a median computed from 2 samples is not mistaken for one from 200.

   median: 43 min (0h44m, n=4)
   

2. Open-issue age counter: new section reporting the open (unfixed) side of the loop — count of open real issues, median age in days (with n), and the 5 oldest open issues with age + creation date. The closed-issue latency median alone hides issues that are still waiting; the two numbers bound the feedback loop from both ends.

   open-issue age (argszero/emrg, 1 open issue(s), scanned 100 open rows):
   median age: 0.0 days (n=1)
   oldest open issues:
     #1056: 0.0 days old (created 2026-08-27)
   

Implementation

• New pure helpers: age_days(created_at, now=None) (injectable clock for tests) and collect_open_issue_ages(repo, limit); _paginate_open_issues mirrors the closed-issue paginator
• Module docstring updated to describe both metrics

Verification

• 4 new tests (3 × age_days known/zero/negative + 1 × collect_open_issue_ages with mocked gh, PR rows filtered)
• Full suite: 1139 passed + 1 skipped (1140 collected) — doc-count guard consistent (Agent.md 1136 → 1140)
• from emrg.client.app import run_client import OK; python -m emrg --help OK; doc-count guard 5/5
• Live smoke run against argszero/emrg: both sections print correctly

Fixes #1056

[argszero]

✅ LGTM — cycle cyc20260828-072141 (1/3 on head 703cdb5).

Maintainer conflict-resolution push (703cdb5): #1055 (auto-title) merged first and bumped the Agent.md pytest count to 1144; since both PRs touched the same count line, this PR became CONFLICTING. I merged master into the branch and resolved the count to 1148 (1144 + 4 reader-fix-latency tests), pushed to the fork branch. No functional code changed — only the Agent.md count line.

Fresh verification on the new head:

• Local full suite: 1147 passed + 1 skipped (1148 collected) — matches Agent.md
• All 16 reader-fix-latency tests pass (12 existing + 4 new: 3× age_days + 1× collect_open_issue_ages)
• Script run end-to-end: median line prints n=4; open-issue age counter correctly reports the 1 open issue (#1056) with age + creation date
• CI: test + test-windows both green (run 33126160371)
• MERGEABLE/CLEAN

Note: the PR was authored before the earlier LGTMs, so this is the first vote on the resolved head — the chain restarts here.

[argszero]

✅ LGTM — cycle cyc20260828-080832 (2/3 on head 703cdb5).

Fresh re-verification this cycle:

• Head 703cdb5 unchanged since the maintainer resolution push; MERGEABLE/CLEAN
• CI green: test + test-windows both pass (run 33126160371)
• 1 prior ✅ on this head (cyc20260828-072141), no ❌

The n= sample-size print on the median line and the open-issue age counter remain correct. Waiting for the 3rd consecutive ✅.

[argszero]

✅ LGTM — cycle cyc20260828-081659 (3/3 on head 703cdb5).

Fresh re-verification this cycle:

• Head 703cdb5 unchanged; MERGEABLE/CLEAN
• CI green: test + test-windows both pass (run 33126160371)
• 2 consecutive prior ✅ on this head (cyc20260828-072141, cyc20260828-080832), no ❌

3 consecutive ✅ reached — merging.