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/PaulJPhilp/EffectPatternsWrote 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/pauljphilp/effectpatterns/implement-dead-letter-queues)<a href="https://agentmods.dev/rules/pauljphilp/effectpatterns/implement-dead-letter-queues"><img src="https://agentmods.dev/badge/rules/pauljphilp/effectpatterns/implement-dead-letter-queues.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 | $0.01548 | $0.01548 |
| Opus 5 | $0.00774 | $0.00774 |
| Sonnet 5 | $0.00310 | $0.00310 |
| Haiku 4.5 | $0.00155 | $0.00155 |
Grade A, and why
implement-dead-letter-queues 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 — 228 lines — stays where its author put it; the contents beside it link to each section on GitHub.
description: Capture failed items with context for debugging and retry instead of losing them. globs: "**/*.ts" alwaysApply: true
Implement Dead Letter Queues
Rule: Capture failed items with context for debugging and retry instead of losing them.
Example
import { Effect, Stream, Queue, Chunk, Ref, Data } from "effect"
// ============================================
// 1. Define DLQ types
// ============================================
interface DeadLetterItem<T> {
readonly item: T
readonly error: unknown
readonly timestamp: Date
readonly attempts: number
readonly context: Record<string, unknown>
}
interface ProcessingResult<T, R> {
readonly _tag: "Success" | "Failure"
}
class Success<T, R> implements ProcessingResult<T, R> {
readonly _tag = "Success"
constructor(
readonly item: T,
readonly result: R
) {}
}
class Failure<T> implements ProcessingResult<T, never> {
readonly _tag = "Failure"
constructor(
readonly item: T,
readonly error: unknown,
readonly attempts: number
) {}
}
// ============================================
// 2. Create a DLQ service
// ============================================
const makeDLQ = <T>() =>
Effect.gen(function* () {
const queue = yield* Queue.unbounded<DeadLetterItem<T>>()
const countRef = yield* Ref.make(0)
return {
send: (item: T, error: unknown, attempts: number, context: Record<string, unknown> = {}) =>
Effect.gen(function* () {
yield* Queue.offer(queue, {
item,
error,
timestamp: new Date(),
attempts,
context,
})
yield* Ref.update(countRef, (n) => n + 1)
yield* Effect.log(`DLQ: Added item (total: ${(yield* Ref.get(countRef))})`)
}),
getAll: () =>
Effect.gen(function* () {
const items: DeadLetterItem<T>[] = []
while (!(yield* Queue.isEmpty(queue))) {
const item = yield* Queue.poll(queue)
if (item._tag === "Some") {
items.push(item.value)
}
}
return items
}),
count: () => Ref.get(countRef),
queue,
}
})
// ============================================
// 3. Process with DLQ
// ============================================
interface Order {
id: string
amount: number
}
const processOrder = (order: Order): Effect.Effect<string, Error> =>
Effect.gen(function* () {
// Simulate random failures
if (order.amount < 0) {
return yield* Effect.fail(new Error("Invalid amount"))
}
if (order.id === "fail") {
return yield* Effect.fail(new Error("Processing failed"))
}
yield* Effect.sleep("10 millis")
return `Processed order ${order.id}: $${order.amount}`
})
const processWithRetryAndDLQ = (
orders: Stream.Stream<Order>,
maxRetries: number = 3
) =>
Effect.gen(function* () {
const dlq = yield* makeDLQ<Order>()
const results = yield* orders.pipe(
Stream.mapEffect((order) =>
Effect.gen(function* () {
let lastError: unknown
for (let attempt = 1; attempt <= maxRetries; attempt++) {
const result = yield* processOrder(order).pipe(
Effect.map((r) => new Success(order, r)),
Effect.catchAll((error) =>
Effect.gen(function* () {
yield* Effect.log(`Attempt ${attempt}/${maxRetries} failed for ${order.id}`)
lastError = error
if (attempt < maxRetries) {
yield* Effect.sleep("100 millis") // Backoff
}
return new Failure(order, error, attempt) as ProcessingResult<Order, string>
})
)
)
if (result._tag === "Success") {
return result
}
}
// All retries exhausted - send to DLQ
yield* dlq.send(order, lastError, maxRetries, { orderId: order.id })
return new Failure(order, lastError, maxRetries)
})
),
Stream.runCollect
)
const successful = Chunk.filter(results, (r): r is Success<Order, string> => r._tag === "Success")
const failed = Chunk.filter(results, (r): r is Failure<Order> => r._tag === "Failure")
yield* Effect.log(`\nResults: ${Chunk.size(successful)} success, ${Chunk.size(failed)} failed`)
// Get DLQ contents
const dlqItems = yield* dlq.getAll()
if (dlqItems.length > 0) {
yield* Effect.log("\n=== Dead Letter Queue Contents ===")
for (const item of dlqItems) {
yield* Effect.log(
`- Order ${item.item.id}: ${item.error} (attempts: ${item.attempts})`
)
}
}
return { successful, failed, dlqItems }
})
// ============================================
// 4. DLQ reprocessing
// ============================================
const reprocessDLQ = <T>(
dlqItems: DeadLetterItem<T>[],
processor: (item: T) => Effect.Effect<void, Error>
) =>
Effect.gen(function* () {
yield* Effect.log(`Reprocessing ${dlqItems.length} DLQ items...`)
for (const dlqItem of dlqItems) {
const result = yield* processor(dlqItem.item).pipe(
Effect.map(() => "success" as const),
Effect.catchAll(() => Effect.succeed("failed" as const))
)
yield* Effect.log(
`Reprocess ${JSON.stringify(dlqItem.item)}: ${result}`
)
}
})
// ============================================
// 5. Run example
// ============================================
const program = Effect.gen(function* () {
const orders: Order[] = [
{ id: "1", amount: 100 },
{ id: "2", amount: 200 },
{ id: "fail", amount: 50 }, // Will fail all retries
{ id: "3", amount: 300 },
{ id: "4", amount: -10 }, // Invalid amount
{ id: "5", amount: 150 },
]
yield* Effect.log("=== Processing Orders ===\n")
const { dlqItems } = yield* processWithRetryAndDLQ(Stream.fromIterable(orders), 3)
if (dlqItems.length > 0) {
yield* Effect.log("\n=== Attempting DLQ Reprocessing ===")
yield* reprocessDLQ(dlqItems, (order) =>
Effect.gen(function* () {
yield* Effect.log(`Manual fix for order ${order.id}`)
})
)
}
})
Effect.runPromise(program)
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 · 228 lines · 1,548 tokens per session scan A 5daee5ac21ef
implement-dead-letter-queues is a cursor rule published in the GitHub repository PaulJPhilp/EffectPatterns (796 stars, last pushed 2mo ago), licensed MIT. It adds 1,548 tokens to every session, about $0.0077 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 2026-09-03.
Other cursor rules, from other repositories
workflow
This file provides rules and context for generating or understanding Go code related to a custom Domain Specific Language (DSL) for defining Temporal workflows within this project.
activities
This file provides rules and context for generating or understanding Go code related to Temporal activities within this project with the specific purpose of using a simple DSL to specify workflows.
go-api-development-general-rules
General rules for Go API development using the net/http package, focusing on code quality, security, and best practices.
backend-general-expert
General rule for backend development expertise across the project.
service-class-conventions
Defines the structure and implementation of service classes, enforcing the use of interfaces, ServiceImpl classes, DTOs for data transfer, and transactional management.
repository-class-conventions
Governs the structure and functionality of repository classes, emphasizing the use of JpaRepository, JPQL queries, and EntityGraphs to prevent N+1 problems.