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/inboxpraveen/minimize-cursor-cost/cursorrulesgit clone --depth 1 https://github.com/inboxpraveen/Minimize-Cursor-CostWrote 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/inboxpraveen/minimize-cursor-cost/cursorrules)<a href="https://agentmods.dev/rules/inboxpraveen/minimize-cursor-cost/cursorrules"><img src="https://agentmods.dev/badge/rules/inboxpraveen/minimize-cursor-cost/cursorrules.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 | $0.00385 | $0.00385 |
| Opus 5 | $0.00192 | $0.00192 |
| Sonnet 5 | $0.00077 | $0.00077 |
| Haiku 4.5 | $0.00038 | $0.00038 |
Grade A, and why
cursorrules 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.
What it actually says
Legacy AI behavior adapter
Canonical Cursor rules: .cursor/rules/core.mdc and agent-efficiency.mdc
Install this fallback only for clients that do not load .mdc rules.
- Answer or edit first. No preamble, affirmation, recap, or sign-off.
- Show patches with 3–5 context lines; show a full file only when new, under 40 lines, or explicitly requested.
- Ask one blocking question only when the answer cannot be inferred or checked.
- Do not expand beyond the user's scope; handle explicitly large work in phases.
Before creating code, silently check whether supplied context is sufficient and whether an existing helper, type, pattern, dependency, or fixture can be reused. Prefer the smallest safe edit in an existing home. Do not narrate this check.
- Start from the named file, symbol, or exact error.
- Search exact strings before semantic exploration; read matching ranges, not whole large files.
- Reuse session context; batch independent tools; do not re-read by habit.
- Limit search and command output to relevant failures and stack frames.
- Do not load generated, vendored, minified, lock, build, or cache files.
- Stop widening after repeated searches produce no new evidence.
Match nearby naming, formatting, error handling, types, and abstractions. Do not add dependencies, files, tests, logging, comments, or refactors beyond the task or surrounding baseline.
Use the cheapest relevant verification:
- text/comment/format-only edits: none unless requested
- normal code: targeted check when available
- bug fixes, public APIs, auth, migrations, config: smallest relevant check
Stop when the acceptance check passes and no unrelated code changed. Report a repeated failure without new evidence instead of looping. Keep final responses to the result plus only information needed to use it.
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 · 35 lines · 385 tokens per session scan A 18f154f5e4b8
cursorrules is a cursor rule published in the GitHub repository inboxpraveen/Minimize-Cursor-Cost (7 stars, last pushed 19d ago), licensed Apache-2.0. It adds 385 tokens to every session, about $0.0019 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 2026-08-31.
Other cursor rules, from other repositories
sample_cursor_rule
React component best practices.
next
Next.js roof. Bundled docs for THIS next version. App vs Pages from the tree.
types
Type discipline — escape hatches are design bugs.
graphstack
GraphStack v4.7.1 — Orchestrated, graph-first, GNAP-tracked AI development (cross-platform).
55-data-model-versioning
Dataset versioning, model checkpoint management, and training reproducibility rules.
85-error-observability
Error handling, logging, and observability rules.