What is swift-ai-sdk AGENTS.md?

A project guide for Swift AI SDK, a Swift Package Manager library that aims to match the behavior of Vercel’s TypeScript AI SDK. It describes the source of truth, repository layout, upstream tracking, APIs, streaming, providers, errors, tools, and tests.

Use it when implementing or checking Swift AI SDK features, matching request and response behavior, handling streaming or providers, updating tests, or comparing work with the pinned TypeScript source.

It helps an agent preserve compatibility with the TypeScript library when changing the Swift version. It also clarifies which files contain verified progress and which upstream files are read-only references.

How teunlao/swift-ai-sdk configures Codex and OpenCode

Reuse

Borrowing it

Nothing to install: this file belongs to teunlao/swift-ai-sdk. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/teunlao/swift-ai-sdk/main/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/teunlao/swift-ai-sdk

Made for: Codex, OpenCode.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/teunlao/swift-ai-sdk/agents-md/github.svg)](https://agentmods.dev/instructions/teunlao/swift-ai-sdk/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/teunlao/swift-ai-sdk/agents-md"><img src="https://agentmods.dev/badge/instructions/teunlao/swift-ai-sdk/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/teunlao/swift-ai-sdk/agents-md"><img src="https://agentmods.dev/badge/instructions/teunlao/swift-ai-sdk/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 1,245 This file is loaded in full into every session.

When invoked 1,245 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.01245	$0.01245
Opus 5.5	$0.00498	$0.00498
Sonnet 5.5	$0.00249	$0.00249
Haiku 4.5	$0.00125	$0.00125

Measured 20d ago against content hash c762b0ea616e, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

swift-ai-sdk AGENTS.md 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 20d 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.

AGENTS.md · 113 lines

How it starts

The opening of the file, as written. The whole thing — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Swift AI SDK

SwiftPM port of the Vercel AI SDK (ai, @ai-sdk/*). The goal is upstream behavior parity: public APIs, request shapes, response parsing, streaming order, errors, tools, usage, and defaults must match the TypeScript source unless a deliberate Swift contract says otherwise.

Reply to the user in Russian. Keep this file in English.

Commands

swift build
swift test                                    # full suite
swift test --filter OpenAIProviderTests       # single target
node tools/test-runner.js --config tools/test-runner.default.config.json
pnpm run docs:check && pnpm run docs:build    # required before a docs page is done
pnpm run example <name>                       # examples/ is a separate package

Use tools/test-runner.js when a suite stalls: it enforces a per-test timeout, reaps zombie swiftpm-testing-helper processes, and prints only the summary and failures. Plain swift test just hangs.

CI (.github/workflows/test.yml) runs swift test --parallel on macOS only, so a race that only shows up under parallel execution will fail there, not locally.

Layout

Sources/AISDKProvider — contracts mirroring @ai-sdk/provider. YOU MUST keep this target dependency-free.
Sources/AISDKProviderUtils — shared HTTP, JSON, schema, validation, retry, header, media, and tool helpers (@ai-sdk/provider-utils).
Sources/SwiftAISDK — the ai package: generate/stream, object generation, tools, agents, registry, middleware, telemetry, UI message streams. All high-level orchestration belongs here, never in a provider target.
Sources/<Name>Provider — one target per provider. Provider-specific request building, auth, streaming, warnings, and decoding stay inside it.
upstream/UPSTREAM.md — pinned upstream commit.
apps/docs — Starlight docs site. examples/ — separate SwiftPM package.
external/vercel-ai-sdk — read-only upstream checkout. Nothing in Sources, Tests, docs builds, or scripts may depend on it, fixtures included.

Read the full file on GitHub · 113 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.

20d ago Changed · -14 lines · -174 tokens per session c762b0ea616e
22d ago Changed · -129 lines · -1,119 tokens per session d3ef52b3be0e
2mo ago First seen · 256 lines · 2,538 tokens per session scan A a4cde5861ae9

Subscribe to this mod's changes

swift-ai-sdk AGENTS.md is an instructions file published in the GitHub repository teunlao/swift-ai-sdk (156 stars, last pushed 20d ago), licensed Apache-2.0. It adds 1,245 tokens to every session, about $0.0050 per session on Opus 5.5. 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 2026-08-30.

Kotlin and Swift Languages ai claude openai sdk swift vercel

swift-ai-sdk AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
swift-ai-sdk AGENTS.mdteunlao/swift-ai-sdk	Instructions file	156	20d ago	Apache-2.0	A	1,245	Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

swift-ai-sdk: Instructions file for Codex