AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Acp Dev Patterns

skill-orcaqubits-agentic-commerce-skills-plugins-acp-dev-patterns · by OrcaQubits

Cross-cutting ACP development patterns — idempotency, error handling, 3D Secure flows, request signing, rate limiting, monitoring, and security best practices. Use when designing architecture or solving production concerns.

No reviews yet
0 installs
26 views
0.0% view→install

Install

$ agentstack add skill-orcaqubits-agentic-commerce-skills-plugins-acp-dev-patterns

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-orcaqubits-agentic-commerce-skills-plugins-acp-dev-patterns)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
4mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Acp Dev Patterns? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

ACP Development Patterns

Before writing code

Fetch live docs:

  • Checkout spec: Fetch https://developers.openai.com/commerce/specs/checkout/ for error codes and idempotency rules
  • Production guide: Fetch https://developers.openai.com/commerce/guides/production/ for operational requirements
  • Stripe integration: Web-search site:docs.stripe.com agentic-commerce for PSP-side patterns
  • Changelog: Web-search site:github.com agentic-commerce-protocol CHANGELOG for latest changes

Pattern: Idempotency

Every POST request MUST include Idempotency-Key (UUID v4, max 255 chars).

Server-side implementation:

  • Store key → response mapping for minimum 24 hours
  • Same key + identical body = replay original response, add Idempotent-Replayed: true header
  • Same key + different body = 422 idempotency_conflict
  • Key in flight = 409 idempotency_in_flight with Retry-After header
  • 5xx responses are NOT cached — retries treated as fresh requests

Client-side implementation:

  • Generate UUID v4 for each logical operation
  • Reuse the same key when retrying a failed request
  • Generate a new key for a genuinely new operation
  • Handle 409 by waiting per Retry-After then retrying

Pattern: Error Handling

ACP uses a flat error structure:

{type, code, message, param}
  • type: invalid_request | processing_error | service_unavailable
  • code: Well-known identifier (e.g., idempotency_conflict, invalid_card, rate_limit_exceeded)
  • message: Human-readable description
  • param: JSONPath (RFC 9535) to the offending field

Retry strategy:

  • 429 rate_limit_exceeded — Exponential backoff with jitter
  • 409 idempotency_in_flight — Wait per Retry-After header
  • 5xx — Retry with same idempotency key, exponential backoff
  • 4xx (except 429) — Do not retry, fix the request

Pattern: 3D Secure Authentication

When complete returns authentication_required:

  1. Extract 3DS challenge from the response
  2. Present challenge to the buyer (redirect or modal)
  3. Buyer completes authentication
  4. Call complete again with authentication_result:
  • three_ds_cryptogram
  • electronic_commerce_indicator
  • transaction_id
  • version
  1. Merchant processes the authenticated payment

Handle authentication timeouts and failures gracefully.

Pattern: Request Signing

For request integrity verification:

  • Signature header: Base64-encoded signature over canonical JSON body
  • Timestamp header: RFC 3339 for freshness validation
  • Algorithm: HMAC (Base64-encoded HMAC signature over canonical JSON body)
  • Verification: Reconstruct canonical JSON, verify signature, check timestamp within clock-skew window

Pattern: Webhook Delivery

Merchant → Agent webhook best practices:

  • Sign EVERY event with HMAC-SHA256
  • Include event ID for consumer-side deduplication
  • Implement exponential backoff for failed deliveries
  • Set reasonable timeouts (5-30 seconds)
  • Log delivery attempts and responses
  • Queue events if the webhook endpoint is down
  • Verify the receiving endpoint uses HTTPS

Pattern: API Versioning

  • Always send API-Version: YYYY-MM-DD header
  • Pin to a specific version in your code
  • Test against new versions before upgrading
  • Handle version-specific response shape differences
  • Log the version used for debugging

Pattern: Rate Limiting

  • Implement server-side rate limiting per API key
  • Return 429 with Retry-After header when exceeded
  • Client-side: Honor Retry-After, add jitter to prevent thundering herd
  • Monitor rate limit proximity and alert before hitting limits

Pattern: Monetary Amount Safety

  • ALL amounts as integers in minor currency units
  • $19.99 = 1999 (cents)
  • Never use floating-point for money
  • Validate on input and output
  • Use language-specific money libraries (e.g., decimal in Python, BigInt in JS)

Pattern: Monitoring & Observability

  • Log every checkout operation with session ID and status
  • Track conversion funnel: create → update → complete → order
  • Alert on error rate spikes (especially 5xx)
  • Monitor idempotency key reuse patterns
  • Track payment success/failure rates
  • Dashboard SPT provisioning and consumption
  • Monitor webhook delivery success rates

Pattern: Security Checklist

  • TLS 1.2+ on all endpoints
  • Bearer token rotation strategy
  • No PCI data in logs (redact card numbers, CVCs)
  • IP allowlisting for agent platform connections
  • Request signing for integrity verification
  • HMAC verification on all incoming webhooks
  • Rate limiting to prevent abuse
  • Input validation on all fields
  • CORS configuration (if browser-facing)

Fetch the latest spec and production guide for current error codes, header requirements, and security recommendations before implementing.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.