Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/proggarapsody/bitbottle/06-tokenomics-speedgit clone --depth 1 https://github.com/proggarapsody/bitbottleWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/proggarapsody/bitbottle/06-tokenomics-speed)<a href="https://agentmods.dev/agents/proggarapsody/bitbottle/06-tokenomics-speed"><img src="https://agentmods.dev/badge/agents/proggarapsody/bitbottle/06-tokenomics-speed.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.04062 |
| Opus 5 | $0.00000 | $0.02031 |
| Sonnet 5 | $0.00000 | $0.00812 |
| Haiku 4.5 | $0.00000 | $0.00406 |
Grade A, and why
06-tokenomics-speed scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tokenomics & Speed — Cycles 158–187
Analysis date: 2026-06-02
Dimension: Tokenomics & Speed
Window: Cycles 153–187 (30 cycles logged); token-valid subset: 157–164 (May-25 era) + 178–187 (Jun-1 stream). Cycles 168–177 log "tokens": 0 — fully excluded as missing instrumentation, not treated as free.
TL;DR
- TDD is the dominant cost: 67.5% of attributed tokens and ~90% of active wall time.
- Design Judge is surprisingly expensive: 40% of
(TDD + DJ)token budget at ~75K tokens per cycle despite running in 83 s median. - 2.7× speed improvement confirmed: May-25 era median 24.5 min → Jun-1 stream median 9.0 min/cycle.
- Token cost rose 3× alongside the speedup: May median 62.5K → Jun median 188K tokens/cycle. Fewer CI wait minutes; more subagent work per minute.
- Rework cycles (shipped_with_fix) cost an extra ~45 min wall time across the stream but show no intrinsic token premium on the primary cycle — the hidden cost is the untracked follow-up fix PRs (~90K tokens each × 5 = ~450K extra tokens not in dataset).
1. Token Distribution
Iteration cycles with real token data (n=14)
| Statistic | Value | Cycle |
|---|---|---|
| Minimum | 31,189 | c158 (REPO-DOWNLOAD, simple scope) |
| Median | 162,076 | — |
| Mean | 141,806 | — |
| Maximum | 235,426 | c186 (ADMIN-RATE-LIMIT) |
| Std dev | 71,338 | — |
Per-cycle breakdown
| Cycle | Era | Outcome | Tokens | Active min |
|---|---|---|---|---|
| 157 | May-25 | shipped | 105,380 | 24.7 |
| 158 | May-25 | shipped | 31,189 | 24.5 |
| 159 | May-25 | shipped | 32,570 | 24.1 |
| 160 | May-25 | shipped | 62,547 | 22.6 |
| 164 | May-25 | shipped | 172,481 | 41.1 |
| 178 | Jun-1 | shipped | 162,816 | 9.0 |
| 179 | Jun-1 | shipped (chore) | 55,860 | 1.9 |
| 180 | Jun-1 | shipped_with_fix | 218,592 | 11.6 |
| 181 | Jun-1 | shipped | 220,483 | 13.9 |
| 183 | Jun-1 | shipped_with_fix | 150,322 | 6.4 |
| 184 | Jun-1 | shipped_with_fix | 161,336 | 7.2 |
| 185 | Jun-1 | shipped_with_fix | 190,742 | 9.4 |
| 186 | Jun-1 | shipped_with_fix | 235,426 | 9.9 |
| 187 | Jun-1 | shipped_with_fix | 185,542 | 8.7 |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 255 lines · 0 tokens per session scan A e8c2f853ceab
06-tokenomics-speed is an agent published in the GitHub repository proggarapsody/bitbottle (3 stars, last pushed 17d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 4,062 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
release-bundle-testing
Rozenite is a development tool. Whatever a project wires into its bundler config, a release build must ship none of our code. @rozenite/test-utils provides the bench that proves it, and every plugin owns a Vitest suite in src/tests/release-bundle.test.ts that uses it.
e2e-testing
pnpm expo run:ios pnpm expo run:android.
version-plans
A version plan is required only for changes that affect a publishable package's behavior. Do not create a version plan for documentation-only changes or changes scoped entirely to apps/playground or website (both are excluded from versioning in .changeset/config.json).
playground-testing
Rozenite DevTools plugins. Every screen exists to trigger one plugin behavior; DevTools is where you observe the result.
debugger
Debugs errors, test failures, and unexpected behavior. Knows PromptScript architecture.
domain
This document defines the canonical domain language for Vibelution so architecture work can use stable names for the same concepts. It is intentionally small: add terms when they become load-bearing in design or tests.