diff --git a/Reviews/2026-year-in-runs.md b/Reviews/2026-year-in-runs.md
new file mode 100644
index 0000000..3b1c52f
--- /dev/null
+++ b/Reviews/2026-year-in-runs.md
@@ -0,0 +1,92 @@
+---
+date: 2026-09-13
+type: review
+tags: [review, running, year, honest-review]
+---
+
+# 2026 — the year in runs (Jun 7 → Sep 13)
+
+**One line:** the engine got a lot better (+9 VO₂Max, longest run 4.9 → 6.56 km), the tissue did not, and August — 3 runs in 31 days — is the single most expensive event of the year.
+
+## 1. The volume picture
+
+| Month | Runs | Distance | Avg / run | Note |
+|-------|------|----------|-----------|------|
+| Jun (7–30) | 9 | 25.45 km | 2.83 km | start of the log |
+| Jul (1–29) | 13 | 42.09 km | 3.24 km | peak month, PBs, festival |
+| **Aug** | **3** | **11.43 km** | **3.81 km** | **two ~2-week holes** |
+| Sep (2–13) | 6 (1 aborted) | 23.41 km | 3.90 km | highest load rate of the year |
+| **Total** | **31** | **~102.4 km** | ~3.3 km | ~2.2 runs/week |
+
+Gaps (days with zero running):
+
+| Gap | Length | Reason |
+|-----|--------|--------|
+| Jun 22 → Jun 29 | 7 days | unlogged |
+| Jul 29 → Aug 11 | 13 days | heat week → vacation |
+| **Aug 15 → Aug 29** | **14 days** | heat week → vacation → sick week |
+| Sep 10 → Sep 13 | 3 days | peroneal flare |
+
+**August in one number: 3 runs in 31 days, a 90% drop from July.** Then he came back with the highest-volume stretch of the entire year (Sep 2–13 ≈ 1.95 km/day ≈ a 59 km/month rate) — after a month off, on the worst-conditioned ankles of the year.
+
+## 2. What genuinely improved (engine)
+
+| Metric | Jun | Now | Delta |
+|--------|-----|-----|-------|
+| VO₂Max (watch) | 37 (Jun 10) | 46 (since Aug 29) | **+9 (+24%)** |
+| Longest run | 4.90 km (Jun 7) | 6.56 km (Aug 29) | +1.66 km |
+| Fastest avg pace | 6'29" (Jun 7/13) | 5'21" /km (Jul 29, 2 km) | — |
+| Best 3 km | — | 17:55 @ 5'57" (Jul 20) | — |
+| Resting HR | — | 48–50 bpm | — |
+| HRV top | — | 145 ms (Sep 2, Sep 8) | — |
+| Best sleep score | — | 91/100 (Sep 10) | — |
+
+**The peak pace was not lost.** Same-effort pairs:
+
+| Effort | July | September |
+|--------|------|-----------|
+| ~3 km @ ~157 bpm | Jul 20: 3.01 km, **5'57"** @ 158 | Sep 9: 3.24 km, **5'58"** @ 157 |
+| ~6.5 km @ ~157 bpm | Jul 23: 6.52 km, 6'34" @ 160 | Sep 6: 6.07 km, 6'55" @ 157 |
+
+Short-run speed is flat. **Long-run pace is still 20–35 s/km off July.** "I'm back at level 0" is not what the data says; "my long-run endurance was reset by August" is.
+
+## 3. What did not improve (the honest half)
+
+1. **Consistency.** 2.2 runs/week average, best month 13 runs, worst month 3. He has never had one full calendar month at 3 runs/week. Every gap cost more than the training before it earned.
+2. **Tissue, not engine, is the ceiling.** The 2026 injury chain: right ankle (Jun–Jul) → right inner thigh adductor → right thigh above knee → left achilles (Sep 6) → bilateral calf–achilles tightness (Sep 9/10) → left peroneal flare (Sep 12/13). Root cause found only on Sep 13: **knee-to-wall 4 cm left / 6 cm right vs 10 cm normal.** A 3–4 month project, started in week 14 of the running year.
+3. **Gear.** Kiprun Jogflow 190.1: fine to ~4 km, numbness from 3.5–5 km (Jul 5, Sep 6, Sep 10). The 7 km long run was never shoe-supported; it was scheduled onto a shoe that caps at 4.
+4. **Alcohol + late nights.** Every worst-performance of the year has a drinking night behind it: Jul 8 (rage run, side stitch at 1.5 km), Jul 27 (festival, 150k steps, post-festival blues), Aug 9 (**HRV 29 ms — all-time low**, 4h33 sleep, 5am), Sep 11 (concert → Sep 13 flare). Every top-5 HRV night is sober and pre-midnight. The lever is known and it is not training.
+
+## 4. The August gap — what it actually cost
+
+- 13 + 14 days off, separated by two 2–4 km runs.
+- September's long runs are 20–35 s/km slower than Jul 23 at comparable HR; the watch's first run back (Aug 29) showed 29 anaerobic minutes at a 7'08" recovery pace — the post-illness signature, not a new baseline.
+- Rebuilding came back **too fast for the tissue**: Sep 2 → Sep 13 is the highest-volume stretch of the year, and by the third run (Sep 6) calf–achilles tightness was logged, escalating to bilateral (Sep 9/10) and a flare (Sep 13).
+- Honest cost estimate: **4–6 weeks of progress erased**, plus the 4 weeks spent since not fully recovering.
+
+## 5. Where he stands today (Sep 13)
+
+- Aborted ~2 km run, left lateral ankle (peroneal), limp on the walk home, pain halved after 15 min sitting = soft tissue.
+- HR was never the limiter (chose a pace that kept it under 150 — correct call).
+- 4th consecutive run with calf/achilles-adjacent symptoms. This was not bad luck; it was load on a known weak link.
+- Smoking: quit Sep 6, day 7. Sleep architecture improving (REMs recovering, deep 39%, score 91).
+
+## 6. What the next 10 weeks need (only three things)
+
+1. **Mobility, 10 min daily, unconditional on running** (`Health/mobility.md`). Left gets extra volume. Target: left stops being the thing that breaks — not a number by Nov 29.
+2. **The right shoe, bought this week.** Stability + 8–12 mm lab drop + wide last. Jogflow is capped.
+3. **Run 3×/week with a boring first km** (7'30"+) and Saturday night sacred. Long run +1 km/week from the current max, with the 10 km checkpoint at Oct 4.
+
+Everything else (speed, ego surges, extra bonus runs) is optional and costs recovery.
+
+## Bottom line
+
+**A runner's engine and a beginner's ankles.** 2026 delivered a real +24% aerobic improvement and the discovery of the actual structural limit. The last four weeks have been roughly net-zero — not from lack of discipline, but because a gap, a drink, and a shoe each cost more than the kilometres earned. Fix the three items above and Nov 29 is live; skip the mobility floor and the plan bends to a 10K, honestly, with no drama.
+
+## Connected
+
+- [[Runs/2026-09-13]] (today, aborted)
+- [[Runs/2026-08-29]] (comeback after the 14-day gap)
+- [[Runs/2026-07-23]] (year's best long run)
+- [[Half-marathon-plan]]
+- [[Health/mobility]]
diff --git a/Runs/2026-09-13.md b/Runs/2026-09-13.md
new file mode 100644
index 0000000..e521837
--- /dev/null
+++ b/Runs/2026-09-13.md
@@ -0,0 +1,45 @@
+---
+date: 2026-09-13
+type: run
+tags: [run, aborted, narrative-only, incident]
+---
+
+# Run — 2026-09-13
+
+**Aborted at ~2 km. Fourth consecutive run with calf/achilles-adjacent symptoms, first one to end the run. Left lateral ankle (peroneal / fibularis) — walked home with a half-leg limp. HR was never the limiter: pace was deliberately slow and stayed under 150 bpm. The stopper was the legs, not the engine.**
+
+> ⚠️ **Narrative-only entry.** No Huawei screenshots were sent for this session, so there is no measured distance, duration, pace, HR, cadence or zone data. Distance is an estimate from the report ("barely past 2 km"). If screenshots arrive, this note gets the real numbers.
+
+## Data
+
+| Metric | Value |
+|--------|-------|
+| Distance | ~2.0 km (narrative estimate) |
+| Avg pace | slow on purpose — "deliberately kept the pace down" |
+| Avg HR | < 150 bpm (never exceeded 150) |
+| Duration | — |
+| Cadence / zones / VO₂Max | — (no screenshot) |
+
+**Context of the preceding 48 h:** Friday night (Sep 11) concert in 0-drop flat shoes — hours of jumping and dancing, alcohol, bed after 3 am. Saturday (Sep 12) heavy fatigue, left outer ankle already sore/loaded, no pain while walking in a different shoe. Saturday night: 9 h+ sleep, HRV 110+.
+
+## How it felt
+
+> Both calves tight at the back first, then the left leg started to feel "not right" — felt like running with a slight limp, so he stopped rather than pushing on. Walking home was possible only limping on half the leg. Left lower-leg muscle very tight; left outer ankle painful. After ~15 min of sitting the tightness cleared and the ankle pain halved, but did not disappear.
+>
+> Subjective read: *"The pain is already half gone, after sitting for 15 mins, but it just shows that I really need to work on my feet flexibility/mobility."*
+
+## Read
+
+- **Soft tissue, not bone.** Pain halving after 15 min of sitting is the classic soft-tissue sign (a fracture does not do that). Lateral malleolus region, behind/under the bump = peroneal tendon.
+- **Load chain:** concert (plyometric leg session in 0-drop shoes) + alcohol + 3 am bedtime + a walk-heavy Saturday → Sunday run. HRV recovered (110+, 9 h sleep) but **HRV measures the nervous system, not tendon tissue.** Good HRV + a run that dies at 2 km at low HR = local tissue is the limiter.
+- **Stopping was correct.** A limp is the one thing you do not run through.
+- **Not a fitness loss.** See [[Reviews/2026-year-in-runs]] — September's short-run speed is flat vs July at the same HR; the limiter is tissue, shoes and load.
+
+**Return-to-running gate:** walking limp-free → then 10 single-leg heel raises pain-free → then run. Nothing before both.
+
+## Connected
+
+- [[Reviews/2026-year-in-runs]]
+- [[Health/mobility]]
+- [[Runs/2026-09-10]] (previous run — third consecutive achilles-tightness report)
+- [[Half-marathon-plan]]