chore: upgrade to upstream google/adk-go v1.5.0 (true merge)

.github/actions/setup/action.yml

@@ -11,13 +11,13 @@ runs:
   steps:
     - name: Set up Go
       id: setup-go
-      uses: actions/setup-go@d35c59abb061a4a6fb18e82ac0862c26744d6ab5 # v5.5.0
+      uses: actions/setup-go@4a3601121dd01d1626a1e23e37211e3254c1c06c # v6.4.0
       with:
         go-version-file: 'go.mod'
         cache: false # wrapper for actions/cache that doesn't support all functionality
 
     - name: Load Go cache
-      uses: actions/cache@5a3ec84eff668545956fd18022155c47e93e2684 # v4.2.3
+      uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
       with:
         path: |
           ~/.cache/go-build

.github/dependabot.yml

@@ -0,0 +1,18 @@
+version: 2
+updates:
+  - package-ecosystem: "gomod"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+      day: "monday"
+      time: "02:00"
+    commit-message:
+      prefix: "chore"
+      include: "scope"
+    labels:
+      - "dependencies"
+    reviewers:
+      - "wolo-lab"
+    allow:
+      - dependency-name: "google.golang.org/genai"
+    open-pull-requests-limit: 10

.github/workflows/go.yml

@@ -7,10 +7,10 @@ name: Go
 
 on:
   push:
-    branches: [ "main" ]
+    branches: [ "main", "v2" ]
 
   pull_request:
-    branches: [ "main" ]
+    branches: [ "main", "v2" ]
 
   workflow_dispatch:
 

.github/workflows/nightly.yml

@@ -22,7 +22,11 @@ jobs:
     runs-on: ubuntu-latest
     steps:
     - name: Run govulncheck
-      uses: golang/govulncheck-action@b625fbe08f3bccbe446d94fbf87fcc875a4f50ee # v1.0.4
+      # NOTE: pinned to master HEAD instead of a tag. The latest release (v1.0.4)
+      # internally uses actions/checkout@v4.1.1 and actions/setup-go@v5.0.0 (Node 20),
+      # which are deprecated. master uses checkout v6.0.2 + setup-go v6.2.0 (Node 24).
+      # TODO: re-pin to a tagged release once upstream tags one past v1.0.4.
+      uses: golang/govulncheck-action@31f7c5463448f83528bd771c2d978d940080c9fd # master, post-v1.0.4 (Node 24)
       with:
            go-version-file: go.mod
            repo-checkout: true

.golangci.yml

@@ -53,6 +53,7 @@ linters:
         # Changes to the default config:
         - -QF1001 # Exclude "Apply De Morgan's law"
         - -QF1008 # Exclude "Omit embedded fields from selector expression"
+        - -SA1019 # Exclude to keep the old a2a compatibility, can remove for ADK v2
 
   exclusions:
     rules:

AGENTS.md

@@ -0,0 +1,149 @@
+# AGENTS.md
+
+Context for AI coding agents (Claude Code, Gemini CLI, Cursor, Copilot, etc.)
+working in the ADK Go repository. Human contributors should start with
+CONTRIBUTING.md.
+
+## Project overview
+
+ADK Go (`google.golang.org/adk`) is an open-source, code-first Go toolkit for
+building, evaluating, and deploying AI agents. It is model-agnostic but
+optimized for Gemini, and is one of three ADK implementations — Go, Python, and
+Java — that share a conceptual model but are independent codebases. Requires
+Go 1.25+.
+
+## Setup & core commands
+
+Run from the repo root. These match what CI enforces (CI also passes `-v`):
+
+- Build:       `go build -mod=readonly ./...`
+- Test:        `go test -race -mod=readonly -count=1 -shuffle=on ./...`
+- Single pkg:  `go test -race ./agent/...`
+- Lint:        `golangci-lint run`   (golangci-lint v2; CI pins v2.3.1; config in `.golangci.yml`)
+- Tidy check:  `go mod tidy -diff`   (must print nothing)
+- Format:      `golangci-lint fmt`   (applies gofumpt + goimports per config)
+
+## Definition of done
+
+A change is complete only when all of these pass locally:
+
+1. `go build` (above) succeeds.
+2. `go test` (above) is green.
+3. `golangci-lint run` reports no findings.
+4. `go mod tidy -diff` prints nothing.
+5. New/changed behavior has tests; a bug fix has a test that reproduces the bug.
+6. Every new Go file starts with the Apache 2.0 license header (enforced by `goheader`).
+
+## Repository layout
+
+- `agent/`     Agent interface + types (`llmagent`, `workflowagents`, `remoteagent`)
+- `runner/`    Execution engine that drives the run loop
+- `model/`     LLM abstraction (`gemini`, `apigee`)
+- `tool/`      Tool/Toolset interface + built-in tools (incl. `skilltoolset/`, `mcptoolset/`)
+- `session/`   Conversation state + events
+- `memory/`, `artifact/`   Long-term memory and file/data services
+- `plugin/`    Cross-cutting lifecycle hooks
+- `server/`    HTTP servers (`adkrest` is primary; `adka2a`, `agentengine`)
+- `cmd/`       CLI (`adkgo`) and server launchers
+- `telemetry/`, `util/`   Public helper packages
+- `internal/`  Private packages — NOT public API; `internal/httprr` is vendored
+- `examples/`  Runnable example agents (quickstart, tools, a2a, skills, …)
+
+## Conventions & idioms
+
+- **Streaming:** agent runs return `iter.Seq2[*session.Event, error]`; consume
+  with `for event, err := range … {}`. Don't collect events into a slice.
+- **Interface-first:** public packages expose interfaces (`Agent`, `Tool`,
+  `Toolset`, `Service`); concrete impls live in sub-packages or `internal/`.
+- **Callbacks over subclassing** (`Before*`/`After*` for Agent/Model/Tool);
+  returning non-nil from a `Before` callback short-circuits execution.
+- **Errors:** wrap with `fmt.Errorf("…: %w", err)`. Tool confirmation uses
+  sentinel errors (e.g. `tool.ErrConfirmationRequired`).
+- Prefer an existing helper over a new one; keep packages small and focused.
+
+## Minimal example
+
+```go
+model, err := gemini.NewModel(ctx, "gemini-2.5-flash",
+    &genai.ClientConfig{APIKey: os.Getenv("GOOGLE_API_KEY")})
+// handle err
+a, err := llmagent.New(llmagent.Config{
+    Name:        "assistant",
+    Model:       model,
+    Instruction: "You are a helpful assistant.",
+    Tools:       []tool.Tool{ /* ... */ },
+})
+// handle err
+r, err := runner.New(runner.Config{
+    AppName:           "my-app",
+    Agent:             a,
+    SessionService:    session.InMemoryService(),
+    AutoCreateSession: true,
+})
+// handle err
+msg := genai.NewContentFromText("Hello", genai.RoleUser)
+for event, err := range r.Run(ctx, userID, sessionID, msg, agent.RunConfig{}) {
+    // handle err; read event.LLMResponse.Content
+}
+```
+
+See `examples/quickstart` for a full runnable program.
+
+## Extending the framework
+
+- **Add a tool:** wrap a Go function with
+  `functiontool.New[Args, Results](cfg, handler)` (Args/Results are structs), or
+  implement the `tool.Tool` interface for full control.
+- **Add a toolset:** implement `tool.Toolset`; its `Tools(ctx)` may return
+  different tools per invocation.
+- **Add an agent type:** follow the `agent/workflowagents/*` packages; construct
+  agents via `llmagent.New` / `agent.New`, not by implementing `agent.Agent`
+  directly.
+- **Add cross-cutting behavior:** register a `plugin.New(plugin.Config{...})`
+  hook (`Before*`/`After*` for run/agent/model/tool) instead of editing the loop.
+
+## Testing
+
+- Tests run **offline by default**: LLM HTTP traffic is replayed from
+  `testdata/*.httprr` via `internal/httprr`. Never add live model or network
+  calls to tests.
+- To (re)record a package's traffic, supply real credentials (e.g.
+  `GOOGLE_API_KEY`) and run `go generate ./<pkg>/...` (it runs
+  `go test -httprecord=…`); commit the updated `testdata/*.httprr`.
+- Prefer table-driven tests; shared helpers live in `internal/testutil`.
+
+## Boundaries
+
+**Always**
+- Run build, tests, lint, and `go mod tidy -diff` before declaring done.
+- Keep PRs small and focused — one concern per PR.
+- Add or update tests for the code you change.
+
+**Ask first**
+- Adding or upgrading a dependency (`go.mod`).
+- Changing a high-fan-in package (`session`, `agent`, `model`, `tool`,
+  `runner`) — prefer additive, backward-compatible changes.
+- Any change to the public API surface, and any breaking change.
+
+**Never**
+- Break the public API — keep changes backward-compatible.
+- Edit vendored code (`internal/httprr`) or commit secrets / API keys.
+- Add tests that make live LLM or network calls.
+
+## PRs & commits
+
+See `CONTRIBUTING.md` for the full process and CLA. Key points for agents:
+most PRs (beyond trivial docs/typos) need a linked issue; include a **Testing
+Plan**; attach logs or screenshots for behavior changes (Runner output / ADK Web).
+
+## Alignment with adk-python
+
+[adk-python](https://github.com/google/adk-python) is the source of truth for
+feature behavior. When porting or validating a feature, check parity with the
+Python implementation.
+
+## Resources
+
+- Docs: https://google.github.io/adk-docs/
+- Examples: `./examples`
+- Java ADK: https://github.com/google/adk-java

CLAUDE.md

@@ -0,0 +1 @@
+See [AGENTS.md](./AGENTS.md) for project context, commands, and contribution guidelines for AI coding agents.

GEMINI.md

@@ -0,0 +1 @@
+See [AGENTS.md](./AGENTS.md) for project context, commands, and contribution guidelines for AI coding agents.