Show the measured savings of Beaver's rules (code, cost, speed) from Ponytail's published benchmark. One-shot display. Use for /beaver-gain, "what does beaver save", "beaver impact".
Installs into .claude/skills of the current project.
Are you the author of Beaver Gain?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/t-celor-beaver-gain)
---
name: beaver-gain
description: >
Show the measured savings of Beaver's rules (code, cost, speed) from
Ponytail's published benchmark. One-shot display. Use for /beaver-gain,
"what does beaver save", "beaver impact".
---
Display this card when invoked. One-shot: change no level and write nothing.
Beaver's rules are Ponytail's. The figures are Ponytail 5's published agentic benchmark: headless Claude Code (Opus 5.5, default effort) on 39 tasks (feature tickets in a FastAPI and React repo, bug fixes, security and privacy cases, small apps), 5 runs each, against the same agent without the rules. 18 tasks have hidden checks for correctness and safety. Each figure is the geometric mean of the per-task medians, measured, not computed from this repo.
```
beaver gain Ponytail benchmark · 39 tasks × 5 runs · Opus 5.5
no rules ████████████████████ 100%
Lines of code █████████··········· 47% ▼ 53%
Output tokens ███████████········· 55% ▼ 45%
Cost ███████████████····· 74% ▼ 26%
Time ████████████········ 59% ▼ 41%
Hidden checks passed 97% (no rules 96%)
Tests where the logic needs one 98% (no rules 68%)
This session: /the-dam (tokens the dam saved, measured)
This repo: /beaver-debt (shortcuts deferred)
/beaver-audit (what is still cuttable)
```
Never print a per-repo figure for the rules ("you saved X lines here"): the longer version was never written, so there is nothing to subtract from. The dam's own savings (compression, pruned tools) are measured, and `/the-dam` shows them.
