retry-patterns

name	retry-patterns
description	Implement retry patterns for resilient distributed systems. Outputs retry strategies, exponential backoff, circuit breaker integration, idempotency requirements, and failure budget design.
argument-hint	["operation type","failure modes","latency budget","idempotency requirements"]
allowed-tools	Read, Write

Retry Patterns

Retries are the first line of defence against transient failures in distributed systems. Done wrong, they amplify failures (thundering herd), cause data corruption (non-idempotent operations), or waste budget (retrying unrecoverable errors). Done right, they make systems self-healing.

Retry Decision Tree

Should I retry?

Is the operation idempotent? ──No──► Never retry (or make it idempotent first)
         │
        Yes
         │
Is the error transient? ──No (4xx, business error)──► Don't retry
         │
        Yes (5xx, timeout, connection error)
         │
Is the retry budget exhausted? ──Yes──► DLQ or alert
         │
        No
         │
Apply backoff → retry

Exponential Backoff with Jitter

import random
import time
from functools import wraps
from typing import Callable, TypeVar, Type

T = TypeVar("T")

def retry(
    max_attempts: int = 3,
    base_delay: float = 1.0,
    max_delay: float = 60.0,
    exponential_base: float = 2.0,
    jitter: bool = True,
    retryable_exceptions: tuple = (Exception,),
    non_retryable_exceptions: tuple = (),
):
    def decorator(func: Callable) -> Callable:
        @wraps(func)
        def wrapper(*args, **kwargs):
            for attempt in range(max_attempts):
                try:
                    return func(*args, **kwargs)
                except non_retryable_exceptions as e:
                    raise  # Never retry these
                except retryable_exceptions as e:
                    if attempt == max_attempts - 1:
                        raise  # Last attempt — propagate
                    
                    delay = min(base_delay * (exponential_base ** attempt), max_delay)
                    if jitter:
                        # Full jitter: random(0, delay) — avoids synchronized retries
                        delay = random.uniform(0, delay)
                    
                    time.sleep(delay)
        return wrapper
    return decorator

# Usage
@retry(
    max_attempts=3,
    base_delay=1.0,
    max_delay=30.0,
    retryable_exceptions=(ConnectionError, TimeoutError),
    non_retryable_exceptions=(ValidationError, AuthenticationError),
)
def call_payment_api(order_id: str, amount: float) -> dict:
    return payment_client.charge(order_id, amount)

Tenacity (Production-Grade Python)

import tenacity
import httpx

@tenacity.retry(
    stop=tenacity.stop_after_attempt(3),
    wait=tenacity.wait_exponential(multiplier=1, min=1, max=30) +
         tenacity.wait_random(0, 2),                 # Jitter added
    retry=tenacity.retry_if_exception_type(
        (httpx.ConnectError, httpx.TimeoutException)
    ) | tenacity.retry_if_result(
        lambda r: r.status_code in [429, 503]         # Also retry these status codes
    ),
    reraise=True,
    before_sleep=tenacity.before_sleep_log(logger, logging.WARNING),
    after=tenacity.after_log(logger, logging.INFO),
)
async def call_external_api(url: str) -> httpx.Response:
    async with httpx.AsyncClient() as client:
        return await client.get(url, timeout=5.0)

Retry with Idempotency Key

import uuid

async def charge_with_retry(order_id: str, amount: float) -> dict:
    """Payment operations must be idempotent — use idempotency key."""
    idempotency_key = f"charge:{order_id}"  # Stable key for this operation
    
    for attempt in range(3):
        try:
            result = await stripe.charge.create(
                amount=amount,
                idempotency_key=idempotency_key,  # Stripe deduplicates by key
            )
            return result
        except stripe.APIConnectionError:
            if attempt == 2: raise
            await asyncio.sleep(2 ** attempt)
        except stripe.IdempotencyError:
            # Same key used differently — bug in caller, don't retry
            raise

Circuit Breaker + Retry Interaction

from circuitbreaker import circuit, CircuitBreakerError

@circuit(failure_threshold=5, recovery_timeout=30, expected_exception=Exception)
@retry(max_attempts=3, base_delay=1.0, retryable_exceptions=(ConnectionError,))
def resilient_call(url: str) -> dict:
    """
    Retry handles transient failures (3 attempts with backoff).
    Circuit breaker opens after 5 failures, preventing further retries
    for 30 seconds while downstream recovers.
    """
    return requests.get(url, timeout=5).json()

# Handle circuit open gracefully
try:
    result = resilient_call("http://service/api")
except CircuitBreakerError:
    # Circuit is open — return cached/degraded response
    result = get_cached_fallback()

Anti-Patterns to Avoid

Anti-Pattern	Problem	Fix
Retrying non-idempotent operations	Duplicate charges, double sends	Make operations idempotent first; then retry
Fixed delay retries	Synchronized retries create thundering herd	Exponential backoff + jitter
Retrying 4xx errors	Client errors don't recover with retries	Only retry transient (5xx, network) errors
Infinite retries	System never gives up; blocks indefinitely	Max attempts + DLQ or circuit breaker
No retry logging	Can't diagnose retry storms in production	Log each retry attempt with delay and reason

10 Rules

1. Never retry a non-idempotent operation — make it idempotent first.
2. Only retry transient errors (5xx, network timeouts) — never retry 4xx errors.
3. Exponential backoff with full jitter — prevents synchronized retry storms.
4. Maximum retry attempts with a clear failure path — DLQ, alert, or degrade.
5. Idempotency keys for external API calls — the retry sends the same key, deduplication happens server-side.
6. Circuit breaker pairs with retry — retry handles transient, circuit breaker handles sustained failure.
7. Log each retry with attempt number, delay, and exception — retries are invisible bugs otherwise.
8. Retry budget: the total time spent retrying must fit within the caller's timeout.
9. Different retry policies for different operations — payment (3 attempts, long backoff) vs health check (1 attempt).
10. Test retry behaviour explicitly — inject failures to verify idempotency and backoff calculation.

---

Deep Reference Playbook

The sections below extend this skill into a complete operating playbook so it can run end-to-end inside Claude Code, CoWork, or any agentic tool without further prompting. Pull only the sections you need for a given engagement.

Inputs the skill must collect

Before producing any output, the skill confirms:

1. Objective — the single decision or artifact the user wants out of this session.
2. Context — system, team, customer, product, or domain the work sits inside.
3. Constraints — time, budget, headcount, regulatory, technical, political.
4. Definition of done — what "good" looks like and who signs it off.
5. Audience — who reads or consumes the output (engineer, exec, customer, regulator).
6. Existing artifacts — prior versions, related docs, dashboards, tickets.
7. Risk appetite — how reversible the decision is and how much ambiguity is acceptable.

If any of these are missing, the skill asks targeted clarifying questions before generating output. It never invents constraints the user did not state.

Operating workflow

The canonical workflow for Retry Patterns runs in five stages. Each stage has an explicit exit criterion so the skill knows when to advance.

Stage 1 — Frame. Restate the problem in one paragraph. Name the decision, the deadline, the stakeholders, and the success metric. Surface assumptions explicitly so they can be challenged.

Stage 2 — Diagnose. Inventory the current state with concrete evidence: metrics, quotes, screenshots, configs, tickets. Separate facts from interpretations. Identify the two or three root causes that explain most of the gap, not the long tail of symptoms.

Stage 3 — Design. Generate at least two viable options. For each option, capture: what changes, who owns it, what it costs, what it unblocks, what it risks, and how it could fail. Recommend one with a written rationale.

Stage 4 — Execute. Convert the chosen option into a sequenced plan: milestones, owners, dependencies, gating checks, communication cadence, and rollback triggers. Anything that cannot be assigned an owner and a date is not yet a plan.

Stage 5 — Validate. Define how success will be measured, when the measurement happens, and what action follows each possible result. Schedule the retrospective before the work starts, not after.

Outputs the skill produces

Depending on the request, the skill returns one or more of:

• A one-page brief suitable for an executive reader.
• A detailed working document for the delivery team.
• A decision record capturing the choice, the alternatives, and the rationale.
• A risk register with probability, impact, owner, and mitigation.
• A sequenced action plan with named owners and explicit due dates.
• A measurement plan tied to the success metric.
• A communication plan for stakeholders inside and outside the team.

Every artifact uses clear headings, short paragraphs, and tables where comparison helps. No filler. No restating the prompt. No hedging language when a recommendation is warranted.

Decision logic and trade-offs

The skill applies the following heuristics when choices are not obvious:

• Prefer reversible decisions taken quickly over irreversible decisions taken slowly.
• Optimise for the constraint that bites first — usually time, attention, or trust, not money.
• Default to the simplest design that meets the stated definition of done; add complexity only when a specific requirement forces it.
• Make the cost of being wrong visible so the reader can judge whether the recommendation is proportionate.
• Name the people, not the roles, when assigning ownership; ambiguous ownership produces ambiguous outcomes.

Anti-patterns the skill refuses to emit

Anti-pattern	Why it fails	What the skill does instead
Generic best-practice list with no context	Reader cannot act on it	Tailors recommendations to the stated constraints
Recommendation without trade-offs	Hides the cost of being wrong	Names the price paid for the recommendation
Plan with no owners or dates	Cannot be executed or tracked	Assigns a named owner and a date to every action
Metrics theatre	Measures activity, not outcome	Ties every metric back to the user or business outcome
Boil-the-ocean scope	Nothing ships	Cuts scope to the smallest valuable slice
Buried recommendation	Reader misses the point	Leads with the recommendation in the first paragraph

Quality bar

The skill self-checks each output against these gates before returning it:

1. Can a busy executive understand the recommendation from the first 150 words?
2. Is every claim either evidenced, labelled as an assumption, or removed?
3. Does every action have an owner and a date?
4. Are the trade-offs of the recommendation stated honestly?
5. Is there a measurable success criterion?
6. Would the author be comfortable defending this artifact in a review meeting?

If any gate fails, the skill rewrites the section before returning it.

Worked micro-example

Context: Implement retry patterns for resilient distributed systems. Outputs retry strategies, exponential backoff, circuit breaker integration, idempotency requirements

Frame: the team needs a defensible recommendation within five working days; the audience is a cross-functional steering group; the cost of delay is higher than the cost of being slightly wrong.

Diagnose: the dominant constraint is decision latency, not analytical depth. Existing data is sufficient for a directional call.

Design: two viable options surfaced. Option A optimises for speed and reversibility. Option B optimises for completeness but slips the deadline by two weeks.

Execute: Option A recommended. Plan sequenced into a two-week sprint with named owners, a mid-point checkpoint, and a clear rollback trigger.

Validate: success measured against a single leading indicator at day 30 and a single lagging indicator at day 90. Retrospective scheduled for day 35.

Cadence and follow-through

A one-shot artifact rarely changes outcomes. The skill recommends a lightweight cadence to keep the work alive:

• Weekly: owner posts a five-line status (done, doing, blocked, risk, ask).
• Fortnightly: steering group reviews leading indicators and unblocks dependencies.
• Monthly: retrospective on what the data is teaching the team; adjust plan accordingly.
• Quarterly: revisit the original objective and decide whether to continue, pivot, or stop.

Closing rules of thumb

1. Lead with the recommendation; supporting analysis follows.
2. Treat every output as a draft that will be challenged; pre-empt the obvious objections.
3. Prefer one strong recommendation over three weak options.
4. When the evidence is thin, say so; do not launder uncertainty as confidence.
5. Optimise for the next decision, not for the perfect document.
6. Make it easy for the reader to disagree with you in a structured way.
7. Ship the artifact; iterate against feedback rather than in private.