@@ -63,7 +63,7 @@ Cursor, Windsurf, Cline, Copilot, Aider: copy the matching rules file from this
No.
**What if I really need the 120-line cache class?**
-You don't. Insist anyway and he'll build it — slowly, correctly, while looking at you.
+You don't. Insist anyway and he'll build it. Slowly. Correctly. While looking at you.
**Does it scale?**
The code you never wrote scales infinitely. Zero bugs, zero CVEs, 100% uptime since forever.
@@ -73,15 +73,13 @@ You know exactly why.
## Numbers
-5 coding tasks, same agent with and without ponytail: **−16% tokens, ~4× faster, 293 → 47 lines.** The 246 lines nobody wrote have never caused an incident.
-
-Then six harder tasks — streaming parsers, atomic file sync, auth, a concurrent money ledger — against a no-skill control and the [caveman](https://github.com/JuliusBrussee/caveman) skill, with adversarial security and concurrency probes:
+Six tasks: streaming log parser, atomic file sync, notification dispatcher, validation engine, auth module, concurrent money ledger. One spec each, one fresh agent per arm, same model. Three arms: no skill, the [caveman](https://github.com/JuliusBrussee/caveman) skill, and ponytail. Every arm passes the same adversarial security and concurrency probes. Then the agreement ends:
-
+
-Everyone passes every probe. Ponytail does it with a third of caveman's code, a sixth of control's, and was ~5× cheaper when the surprise feature request landed (96 lines changed vs 487 for control). Every shortcut is marked in-code with its upgrade path. Data: [benchmarks/](benchmarks/).
+**47% fewer tokens than the no-skill agent. 3× faster. A seventh of the code.** The 3,139 lines nobody wrote have never caused an incident. When a surprise feature request hit two of the tasks, ponytail extended in 96 changed lines; caveman needed 413, the no-skill agent 1,115. Every shortcut ponytail took is marked in the code with a `ponytail:` comment naming its upgrade path. Data: [benchmarks/](benchmarks/).
## License
diff --git a/assets/benchmark-loc.svg b/assets/benchmark-loc.svg
index 5da8e8b..be81d52 100644
--- a/assets/benchmark-loc.svg
+++ b/assets/benchmark-loc.svg
@@ -1,11 +1,11 @@
diff --git a/benchmarks/results/2026-06-12-v4-hardening-vs-caveman.md b/benchmarks/results/2026-06-12-v4-hardening-vs-caveman.md
index bd2196d..49f8d8e 100644
--- a/benchmarks/results/2026-06-12-v4-hardening-vs-caveman.md
+++ b/benchmarks/results/2026-06-12-v4-hardening-vs-caveman.md
@@ -91,6 +91,28 @@ locks at 3× the LOC).
with its ceiling; Caveman's code marks only spec-allowed simulated transports,
and its design trade-offs live in the chat report, invisible to a later reviewer.
+## Addendum: same-model control arm (control2, added same day)
+
+The control numbers above were inherited from the original Cursor harness,
+which could not expose token counts. Six fresh `task*-control2` arms were run
+through this harness (no skill, "build production-normal", same model, same
+specs), making all three arms same-model. Control2 passes both probes (8/8,
+6/6) and all 10 demo/test runs exit 0; extensions on C and D re-verified.
+
+| Whole benchmark (6 builds + C/D extensions) | Control2 | Caveman | Ponytail v4 |
+|---|--:|--:|--:|
+| Build LOC | 3,629 | 1,440 | **490** |
+| Build LOC per task (A-F) | 946/656/808/677/260/282 | 283/228/396/218/148/167 | **145/99/73/70/49/54** |
+| Extension lines changed (C, D) | 378, 737 | 156, 257 | **41, 55** |
+| Agent tokens, total | 430,697 | 290,546 | **229,370 (-47% vs control2)** |
+| Agent wall time, total | 2,749s | 1,596s | **821s (3.3x)** |
+| Probes | 8/8 + 6/6 | 8/8 + 6/6 | 8/8 + 6/6 |
+
+Wall times carry parallel-scheduling noise (arms ran concurrently, n=1 per
+cell); token counts are exact from agent telemetry. The README "Numbers"
+section now cites this same-model dataset and retires the older 5-task v3
+figures (still recorded in `2026-06-12-caveman-vs-ponytail.md`).
+
## Residual (honest notes)
- A's spike stats still use observed-hours-only mean+3σ rather than a
+ 
+