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/madebyaris/cursor-composer-rules/composer-verificationgit clone --depth 1 https://github.com/madebyaris/cursor-composer-rulesWrote 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/madebyaris/cursor-composer-rules/composer-verification)<a href="https://agentmods.dev/rules/madebyaris/cursor-composer-rules/composer-verification"><img src="https://agentmods.dev/badge/rules/madebyaris/cursor-composer-rules/composer-verification.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 | $0.00785 | $0.00785 |
| Opus 5 | $0.00392 | $0.00392 |
| Sonnet 5 | $0.00157 | $0.00157 |
| Haiku 4.5 | $0.00078 | $0.00078 |
Grade A, and why
composer-verification 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 3d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Composer verification
A claim without evidence is a guess wearing a confident voice. This contract names the evidence that backs each kind of claim.
Status vocabulary
Use one of these labels anywhere you would otherwise say "done":
| Label | Meaning |
|---|---|
| verified | The matching check ran and you can name the evidence (command output, file diff, screenshot, citation). |
| implemented but unverified | Code changed, but the matching check could not run. State exactly what is missing (env, secret, network, permission). |
| blocked | Cannot proceed without an external decision, system, or credential. State what unblock looks like. |
Never label work verified if you skipped the matching check. Pick implemented but unverified or blocked instead.
Proportional evidence
Match proof depth to blast radius. Don't run a full E2E suite for a typo fix; don't claim a new auth flow works because the file compiles.
If the acceptance sentence is user-visible ("The user can X and see Y"), proof is that surface — running app, request, or screenshot. A unit test or compile is extra, not a substitute.
| Change | Minimum evidence |
|---|---|
| Comment, doc, rename | Visual inspection; lint/typecheck if cheap |
| Single function tweak (not user-visible) | Targeted unit test, or a focused command that exercises the path |
| API/contract change | Concrete request/response example or contract test against the new shape |
| New scaffold or subsystem | Install succeeds, dev/start runs, primary happy path passes, build succeeds |
| User-visible UI / new screen / new feature path | Screenshot, recorded interaction, or browser-tool verification at the acceptance path — not "looks fine in the diff", not npm test alone |
| Data migration | Dry-run output on representative data + rollback plan |
| Security-sensitive change | At least one negative test (denied path) and a positive test |
Things that are not verification
- Re-reading your own diff and feeling confident.
- A confident user-facing summary with no named check — tone is not evidence.
- "It compiles" for runtime behavior.
- "The function looks correct" for output correctness.
- "The docs say so" without trying it.
- "The previous test still passes" when the new path is what changed.
npm test/ unit tests alone when the acceptance sentence is a screen, flow, or user-visible API.
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.
- 3d ago First seen · 63 lines · 785 tokens per session scan A 404bd7125816
composer-verification is a cursor rule published in the GitHub repository madebyaris/cursor-composer-rules (3 stars, last pushed 15d ago), licensed MIT. It adds 785 tokens to every session, about $0.0039 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
performance
Performance: profiling-first optimization, caching, bottleneck patterns.
security
Security: input validation, auth, data protection.
prisma
Prisma: schema modeling, queries, migrations.
supabase
Supabase: RLS, edge functions, realtime.
documentation
Documentation: code docs, READMEs, ADRs, maintenance.
git-workflow
Git workflow: commits, branches, PRs, history management.