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 rules/golid-ai/golid/iteration-surfacegit clone --depth 1 https://github.com/golid-ai/golidWrote 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/rules/golid-ai/golid/iteration-surface)<a href="https://agentmods.dev/rules/golid-ai/golid/iteration-surface"><img src="https://agentmods.dev/badge/rules/golid-ai/golid/iteration-surface.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 5.1 | $0.00041 | $0.01001 |
| Opus 5 | $0.00020 | $0.00500 |
| Sonnet 5 | $0.00008 | $0.00200 |
| Haiku 4.5 | $0.00004 | $0.00100 |
Grade A, and why
iteration-surface 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 5d 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 — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Iteration Surface
Thesis: When the natural feedback loop on a feature exceeds 30 seconds (deploy required, external service cache, browser screenshot, CDN purge), the highest-leverage move is to BUILD the iteration surface first. The 30 minutes building an admin preview, a side-by-side comparison, or a hot-reloadable render harness pays back 5-10x in iteration speed during the actual work.
Refrain:
Loop cost > 30s → build the surface → THEN iterate
When this applies
- Dynamic image rendering (OG/social previews, PDF exports, badge generators). External unfurler caches mean shipping → seeing the result is a multi-minute round trip.
- Email template work. Send-to-yourself loops are slow and inboxes cache. An admin preview that renders the same template to HTML pays back fast.
- Marketing/landing copy iteration. Production deploy is the usual feedback loop; an admin preview surface using real data lets a non-engineer iterate.
- AI prompt engineering surfaces. Running prompts against real data through a debug panel beats waiting for the next user conversation.
- PDF / receipt / invoice rendering where downstream caches or external mail providers make iteration slow.
- Any third-party API integration where the response shape needs visual inspection but the integration itself caches aggressively.
When this does NOT apply
- T0 edits where the feedback loop is already fast (
vinxi devHMR is the iteration surface). - One-off changes that won't be iterated on.
- Features where the iteration cost is genuinely 30s or under via existing tooling.
The leverage calculation
| Loop cost | Iterations to break even on a 30-min surface build |
|---|---|
| 30s | 60 (probably not worth it) |
| 2 min | 15 (borderline) |
| 5 min | 6 (worth it) |
| 30 min (chat-app unfurl cache) | ~1 (always worth it) |
If the answer is "we'll iterate more than 5 times" AND "each iteration costs more than 2 min," build the surface.
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.
- 5d ago First seen · 110 lines · 41 tokens per session scan A b01b186b1d21
iteration-surface is a cursor rule published in the GitHub repository golid-ai/golid (40 stars, last pushed 2mo ago), licensed MIT. It adds 41 tokens to every session and 1,001 once invoked, about $0.0002 per session on Opus 5. 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 2026-08-30.
Other cursor rules, from other repositories
adapter-features
Database-specific features must be implemented in the specialized adapter only. Base adapters (postgres, mysql, etc.) must remain database-agnostic.
unit-tests-tdd
TDD required for behavior changes; ≥80% package coverage on touched packages; unit-test conventions.
integration-tests
Human-readable integration test requests; helpers vs httptest; suites/ vs per-DB placement.
state-management
Use the following stack. Do not introduce or recommend Redux or React Context for shared/global state.
backend
You are an expert in Go, Gin, Gorm, Gen, Cosy (https://cosy.uozi.org/) with a deep understanding of best practices and performance optimization techniques in these technologies.
core
Core token-efficiency and response discipline rules. Always active.