no-bare-casts

A TypeScript coding rule that limits direct type-casting with the `as` keyword in production code. It requires named helper functions or a code change instead.

In plain words
What is it for?
Use it when writing or reviewing production TypeScript that needs a type assertion, especially when deciding whether to rewrite the code or document an unavoidable cast.
Why use it?
Unexplained casts can hide type errors and make unsafe assumptions hard to review. Named helpers make the exception visible and, where needed, record its reason.

Cursor rule for Codex

Install

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.

agentmods
npx agentmods add rules/prisma/orm/no-bare-casts
Clone the repo
git clone --depth 1 https://github.com/prisma/orm

Made for: Codex.

Per session 0 Nothing until a file matches its globs; then the whole rule loads.
When invoked 842 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00842
Opus 5 $0.00000 $0.00421
Sonnet 5 $0.00000 $0.00168
Haiku 4.5 $0.00000 $0.00084

Measured yesterday against content hash 4164e6249cec, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

no-bare-casts 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 yesterday.

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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

.agents/rules/no-bare-casts.mdc · 58 lines

How it starts

The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.

No bare as casts in production TypeScript

Bare as type casts (value as Foo, x as unknown as Bar) are forbidden in production code. The no-bare-cast Biome plugin (TML-2685) fires on every bare as token in production files. The CI ratchet rejects per-PR cast-count increases. Every bare cast must be replaced by one of the three accepted forms or eliminated by a code rewrite.

Three accepted forms

Form When to use
blindCast<T, "Reason">(value) The value is genuinely opaque or unrelated to the target type. The Reason literal documents the compromise at the call site.
castAs<T>(value) The value already satisfies T; the assertion is purely declarative (widening from a subtype, or providing an annotation the type-checker cannot infer).
as const A different semantic — produces a readonly literal type, not a type assertion. Always fine; the rule does not fire on as const.

Both helpers live in @internal/utils/casts:

import { blindCast, castAs } from '@internal/utils/casts';

Rewrite first; cast as last resort; reviewer validates the justification

When an author reaches for blindCast, the first action is to ask whether the code can be rewritten so the cast becomes unnecessary. Only when no rewrite is feasible does the cast happen, and the Reason literal must articulate the compromise in language a reviewer can evaluate. An unconvincing Reason is the reviewer's signal to push back — and the author's signal to go solve the underlying type-system problem properly. blindCast is not a sanctioned tool to reach for; it is the auditable escape hatch when everything else has been tried.

Rewrite options to exhaust before reaching for blindCast:

  • Tighten an input type so the cast becomes unnecessary.
  • Add a runtime check that narrows via a type predicate (if (isUser(x)) { … }).
  • Restructure a generic so the bound carries the needed information.
  • Use satisfies when the goal is type-checking without coercion — expr satisfies T is unaffected by this rule.
  • Use castAs<T>(value) when the value already satisfies the target type.

Read the full file on GitHub · 58 lines

Changes

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.

  1. yesterday First seen · 58 lines · 0 tokens per session scan A 4164e6249cec

Subscribe to this mod's changes

no-bare-casts is a cursor rule published in the GitHub repository prisma/orm (47,579 stars, last pushed today), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 842 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 2026-08-30.