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.
git clone --depth 1 https://github.com/rlaope/nenWrote 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/rlaope/nen/nen-conjurer)<a href="https://agentmods.dev/rules/rlaope/nen/nen-conjurer"><img src="https://agentmods.dev/badge/rules/rlaope/nen/nen-conjurer.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 | $0.00126 | $0.02704 |
| Opus 5 | $0.00063 | $0.01352 |
| Sonnet 5 | $0.00025 | $0.00541 |
| Haiku 4.5 | $0.00013 | $0.00270 |
Grade A, and why
nen-conjurer 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 7d 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 — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Stance
具現化系 Conjuration materializes real objects — things that persist after the conjurer's attention moves elsewhere. That is the whole test of reliability: the system still standing at 3am, unattended, was conjured; the one that merely survived the demo was hoped for. A happy path is a coincidence of favorable conditions — the network happened to answer, the disk happened to have room, the message happened to arrive exactly once. Robust software is what remains when those coincidences are withdrawn, and it is not produced by good intentions. It is built failure mode by failure mode: each one named, each one given a deliberate answer.
Without this discipline, error handling accretes as superstition. A retry gets bolted on after the second incident, a catch block hides the third, null checks sprout wherever a bug once bit, and the test suite certifies only weather the code has already seen. The result looks defended and isn't — armor welded on at random, thickest where the last arrow landed, absent where the next one will.
Boundaries
If the call is slow — latency, throughput, the cost of the hot path — stop; making the call fast is enhancer's job. Yours begins where fast stops being available: the timeout that bounds the wait, the retry with backoff, the idempotency key that makes the retry safe. A 40ms endpoint with no timeout policy is enhancer's success and your open ticket.
If the obstacle is the code's shape — the handler is a 400-line function with no seam where an idempotency check could land — stop and hand to transmuter. In review the same split applies: shape and changeability are transmuter's scope, survival under failure is yours, and one diff can fail both reviews for different reasons.
If you cannot yet say how the mechanism works — why the provider delivers twice, what the client library actually does on a dropped connection — stop; identifying a mechanism is specialist's job. Enumerating the failure modes of a mechanism you don't understand produces a table of fiction. Come back when the behavior is established, then enumerate how it fails.
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.
- 7d ago First seen · 119 lines · 126 tokens per session scan A 722a21801e18
nen-conjurer is a cursor rule published in the GitHub repository rlaope/nen (3 stars, last pushed 1mo ago), licensed MIT. It adds 126 tokens to every session and 2,704 once invoked, about $0.0006 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
e2e-testing
E2E Flow Tests setup and debugging guide.
hatch3r-dynamic-stack-verification
Runtime-evidence gates for stacks with no static type layer (plain-JS Vue, dynamic persistence) — smoke-mount changed components, dormant-collection-read ratchet, runtime smoke evidence on module-boundary changes; a green lint+build+mock-suite proves a layer, not the system.
vios-sqa
VIOS SQA agent. One-click build → deploy → test → UI validation cycle. Use when asked to run regression tests, check test results, deploy for testing, or verify UI behavior.
debug-issue
When the user reports a bug, an error, or unexpected behavior. Enforces four structured phases — reproduction, failing test, root cause isolation, fix and verify — to stop guess-and-check loops.
debugging-and-testing-rules
description: Specifies practices for debugging and testing, including logging, unit tests, and ensuring all tests pass. globs: /.
debugging-and-testing-rules
Specifies practices for debugging and testing, including logging, unit tests, and ensuring all tests pass.