stale-bot-implementation by indurireddy-TF · Pull Request #554 · google/adk-go

indurireddy-TF · 2026-02-10T13:27:09Z

What this PR does

Introduces a stale issue automation bot implemented in Go.

Applies a clear decision workflow:

Marks issues as stale when maintainers request clarification and no response is received within the threshold.

Removes the stale label when users respond.

Alerts maintainers on silent description edits (ghost edits).

Automatically closes issues that remain stale beyond the close threshold.

Supports concurrent issue processing with rate limiting.

Uses configurable time thresholds and labels via environment variables.

gemini-code-assist · 2026-02-10T13:27:28Z

Summary of Changes

Hello @indurireddy-TF, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request delivers a robust automation solution for managing GitHub issues by introducing a Go-based stale bot. The bot intelligently processes issues, identifying those requiring attention or closure based on predefined activity thresholds and maintainer input. It streamlines issue maintenance, ensuring that discussions remain active and unresolved items are handled efficiently, while also providing alerts for potentially overlooked changes.

Highlights

Stale Issue Automation Bot: Introduced a new Go-based bot to automate the management of stale GitHub issues.
Decision Workflow: Implemented a comprehensive decision tree for the bot to determine issue status (stale, active, closed) based on activity and maintainer interaction.
Silent Edit Detection: Added functionality to detect and alert maintainers when a user silently edits an issue's description.
Configurable Thresholds: Enabled configuration of stale and close time thresholds, as well as other operational parameters, via environment variables.
Concurrent Processing with Rate Limiting: Designed the bot to process issues concurrently while adhering to GitHub API rate limits.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Changelog

contributing/samples/stale-bot-agent/PROMPT_INSTRUCTION.txt
- Added a detailed instruction prompt for the AI agent, outlining the decision tree for issue auditing.
contributing/samples/stale-bot-agent/agent.go
- Implemented core logic for fetching GitHub issue data using GraphQL.
- Developed functions to build a chronological timeline of issue events (comments, edits, labels).
- Added replayHistoryToFindState to determine the current state of an issue based on its history.
- Included helper functions for GitHub API interactions like adding/removing labels, posting comments, and closing issues.
contributing/samples/stale-bot-agent/config.go
- Defined and initialized configuration variables for GitHub API access, repository details, label names, and time thresholds.
- Implemented utility functions to retrieve environment variables with fallbacks for integers, floats, and strings.
contributing/samples/stale-bot-agent/main.go
- Set up the main application entry point for the stale bot.
- Loaded the AI agent's prompt instructions and configured the Gemini model.
- Orchestrated the processing of issues, including fetching old open issues and processing them concurrently.
- Provided logging for the bot's operations and performance metrics.
contributing/samples/stale-bot-agent/util.go
- Implemented HTTP request utilities with retry and backoff logic for robust API calls.
- Included functions for GET, POST, PATCH, and DELETE requests to the GitHub API.
- Added an API call counter for monitoring usage.
- Developed GetOldOpenIssueNumbers to search for issues meeting the staleness criteria.

Ignored Files

Ignored by pattern: .github/workflows/** (1)
- .github/workflows/stale.yml

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces a new stale bot agent for GitHub repositories, implemented in Go using the ADK framework. The agent's core functionality is defined in agent.go, which includes fetching issue data via GraphQL, building a chronological event timeline, and replaying this history to determine an issue's current state (e.g., last actor, activity time, whether it's stale, or if a maintainer alert is needed for silent edits). It also provides tool functions for interacting with GitHub, such as adding/removing labels, posting comments, and closing issues. The PROMPT_INSTRUCTION.txt file outlines the detailed decision-making workflow for the AI agent, covering scenarios for active, stale, and pending issues based on user roles and activity. Configuration parameters like thresholds and API limits are managed in config.go, while main.go orchestrates the agent's execution, processing issues concurrently in chunks and logging performance metrics. Utility functions for robust HTTP requests with retries and API call counting are provided in util.go. Review comments primarily focus on improving error handling and consistency: specifically, using safer type assertions for GraphQL error responses, handling potential json.Marshal errors in HTTP requests, correcting a placeholder format in the prompt instruction, changing a log.Printf level from 'FATAL' to 'ERROR', suggesting a more efficient isMaintainer check using a map, and standardizing variable naming conventions for labels.

verdverm · 2026-02-13T01:37:01Z

Why is an agent and LLM needed for this? Why is a rule based system insufficient? It looks like the rules are well defined and I'm pretty sure this is what most people are doing when I have seen it over the years (since before LLMs)

hyangah · 2026-04-09T21:11:39Z

+      - name: Set up Go
+        uses: actions/setup-go@v4
+        with:
+          go-version: '1.24' # Use a stable Go version


This is unsupported go version.

updated with new 1.26 version #800

hyangah · 2026-04-09T21:13:07Z

+        working-directory: contributing/samples/stale-bot-agent
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          GOOGLE_API_KEY: ${{ secrets.GOOGLE_API_KEY }}


Does it require this project store secrets in github to support this? Currently there is no github workflow that requires extra secret setup. This will be the first that introduces the need so we need good justification on why LLM is necessary.

We actually implemented this exact same LLM-based stale bot in adk-python as a sample, so the main goal here is to maintain consistency across the Go and Python libraries. The LLM is used to build a "Smart" bot that reads comment intent rather than just checking timestamps. #800

hyangah · 2026-04-09T22:36:52Z

+        uses: actions/checkout@v3
+
+      - name: Set up Go
+        uses: actions/setup-go@v4


Pin to the specific commit (following the google internal security guideline for github action)

hyangah · 2026-04-09T22:41:17Z

+name: ADK Stale Issue Auditor (Go)
+
+on:
+  workflow_dispatch:


Is it true that this allows anyone with the write permission to run this workflow on any branch? If so I think anyone with write permission can theoretically run this on custom branch and access the secret. We need to tie this to a GitHub Environment and configure it to be accessible only from the protected branch.

I originally mirrored the .github/workflows/stale.yml exactly from the adk-python repository to ensure consistency across both projects (their workflow currently doesn't use Environment protections either).
However, you are absolutely right about the risk of workflow_dispatch exposing the secret on custom branches.

hyangah · 2026-04-09T22:41:33Z

@@ -0,0 +1,108 @@
+// Copyright 2025 Google LLC


hyangah · 2026-04-09T22:44:11Z

@@ -0,0 +1,314 @@
+// Copyright 2026 Google LLC


Can we have unit tests to verify at least the basic functionalities? (issue parsing, correctly determining staleness, ...)

added necessary testcases #800

hyangah · 2026-04-09T22:51:24Z

+
+func InitConfig() {
+	GitHubToken = os.Getenv("GITHUB_TOKEN")
+	log.Printf("GITHUB_TOKEN length: %d", len(GitHubToken))


is this logging needed?

Not needed ,Removed.#800

indurireddy-TF · 2026-04-29T19:57:14Z

Hi @hyangah Due to a permission error when trying to push updates to this branch, I am going to create a new PR from my fork and include all of these changes there. I will close this one and share the link to the new PR shortly.

indurireddy-TF · 2026-05-06T14:06:34Z

Hi @hyangah #800 This is a new PR with all necessary test cases and new changes. Please take a look at it.

stale-bot-implementation

9dd2795

gemini-code-assist Bot reviewed Feb 10, 2026

View reviewed changes

indurireddy-TF added 4 commits February 11, 2026 07:12

fix-lint-issues

cd9f5c0

fix-lint-issues

e0bee10

fix-lint-issues-1

864ef5f

fix-testcases

6c8af7a

indurireddy-TF requested review from dpasiukevich and mazas-google February 25, 2026 03:42

hyangah reviewed Apr 9, 2026

View reviewed changes

Conversation

indurireddy-TF commented Feb 10, 2026

Uh oh!

gemini-code-assist Bot commented Feb 10, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

verdverm commented Feb 13, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

indurireddy-TF May 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

indurireddy-TF commented Apr 29, 2026

Uh oh!

indurireddy-TF commented May 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

indurireddy-TF May 6, 2026 •

edited

Loading