madhavkishor/OpenAI-based PR Reviewer & Summarizer

OpenAI-based PR Reviewer and Summarizer

View on GitHub

Trust Signals

Scorecard Score: not yet scored
Maintenance Recency: Maintainedlast commit Nov 3, 2025
License: MIT
Runtime: Deprecated runtime

Pinned Snippet

workflow.ymlSHA-pinned

uses: madhavkishor/openai-pr-reviewer@d7e874900bb555472ffddbfd41648bac4e5b68fa # no releases — HEAD as of 2026-07-11

tags can be moved; commit SHAs can't. why a SHA?

Inputs

name	description	required	default
debug	Enable debug mode	no	false
max_files	Max files to review. Less than or equal to 0 means no limit.	no	60
review_comment_lgtm	Leave comments even if the patch is LGTM	no	false
path_filters	The path filters, e.g., "src/.py", "!dist/", each line will be considered as one pattern. See also - https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#onpushpull_requestpull_request_targetpathspaths-ignore - https://github.com/isaacs/minimatch	no	!dist/ !/.pb.go !/.lock !*/.yaml !*/.yml !*/.cfg !*/.toml !*/.ini !*/.mod !*/.sum !*/.json !*/.mmd !*/.svg !*/.png !*/.dot !*/.md5sum !/gen/ !/_gen/ !/vendor/
openai_model	Model to use	no	gpt-3.5-turbo
openai_model_temperature	Temperature for GPT model	no	0.0
openai_retries	How many times to retry openai API in case of timeouts or errors?	no	5
openai_timeout_ms	Timeout for openai API call in millis	no	60000
openai_concurrency_limit	How many concurrent API calls to make to openai servers?	no	4
system_message	System message to be sent to OpenAI	no	You are `@openai`, a highly experienced software engineer with a strong ability to review code changes thoroughly.You can uncover issues such as logic errors, syntax errors, out of bound errors, potential data races, livelocks, starvation, suspension, order violation, atomicity violation, consistency, complexity, error handling, typos, and more. You will be conducting code reviews today and write code if asked to do so.
summarize_beginning	The prompt for the whole pull request	no	$system_message In this session, we will summarize a pull request. The pull request has the title "$title" and the following description: ``` $description ``` Reply "OK" to confirm that you are ready.
summarize_file_diff	The prompt for each file diff	no	I am providing diff for `$filename` below. I would like you to summarize the diff within 30 words. ```diff $file_diff ```
summarize	The prompt for final summarization response	no	This is the end of the summarization session. Below is the summary you have generated so far for each file. ``` $summary ``` Please provide your final response in the `markdown` format with the following content: - Thank the user for letting you participate in the code review. - High-level summary (focus on the purpose and intent within 80 words) - Table of files and their summaries. You can group files with similar changes together into one row to save space. Avoid additional commentary as this summary will be added as a comment on the pull request.
summarize_release_notes	The prompt for generating release notes	no	Create concise release notes in `markdown` format for this pull request, focusing on its purpose and user story. In needed, you can classify the changes as "New Feature", "Bug fix", "Documentation", "Refactor", "Style", "Test", "Chore", "Revert", and provide a bullet point list. For example: "New Feature: An integrations page was added to the UI". Keep your response within 50-100 words. Avoid additional commentary as this response will be used as is in our release notes.
review_beginning	The beginning prompt of a code review dialog	no	$system_message I will provide you with some files and the entire diff to help you build context, unless the content is too large. Then, I will send you each patch from the diff for review. The pull request has the title "$title" and the following description: ``` $description ``` The OpenAI-generated summary is as follows: ``` $summary ``` Reply "OK" to confirm.
review_file	The prompt for each file	no	Here is the content of `$filename` for context. Please use this context when reviewing patches. ``` $file_content ``` Reply "LGTM!" if the existing code in this file looks correct. If there are serious issues in the existing code please let me know and I will add your comment on the pull request for this file.
review_file_diff	The prompt for each file diff	no	Here is the entire diff for `$filename` for context. Please use this context when reviewing patches. ```diff $file_diff ``` Reply "OK" to confirm.
review_patch_begin	The prompt for each file diff	no	Next, I will send you a series of patches. Each patch consists of a diff snippet, and your task is to review every patch. Identify any bug risks or provide improvement suggestions. If the patch looks good and acceptable, please reply "LGTM!" with a short comment. Your responses will be recorded as review comments on the pull request. Markdown format is preferred for your responses. Reply "OK" to confirm.
review_patch	The prompt for each chunks/patches	no	$filename Existing comments on the patch, though some of them may be outdated. Please use them as additional context. ``` $comment_chain ``` Diff for review: ```diff $patch ```
comment_beginning	Prompt for comment	no	$system_message A comment was made on a review for a diff patch on file `$filename`. I would like you to follow the instructions in that comment. If possible, I will provide you the file and the entire diff to help provide overall context for your response. The pull request has the title "$title" and the following description: ``` $description ``` The OpenAI-generated summary is as follows: ``` $summary ``` Reply "OK" to confirm.
comment_file	Prompt for file	no	Here is the content of `$filename` for context. ``` $file_content ``` Reply "OK" to confirm.
comment_file_diff	Prompt for file diff	no	Here is the entire diff for `$filename` for context. ```diff $file_diff ``` Reply "OK" to confirm.
comment	Prompt for comment	no	I would like you to follow the instructions in the new comment made on a conversation chain on a code review diff. Diff being commented on: ```diff $diff ``` The format of the conversation chain is: `user: comment` Conversation chain (including the new comment): ``` $comment_chain ``` Please reply directly to the new comment (instead of suggesting a reply) and your reply will be posted as-is. If the comment contains instructions/request for you, please comply. For example, if the comment is asking you to generate documentation comments on the code, in your reply please generate the required code. In your reply, please make sure to begin the reply by tagging the user with "@user". The comment/request that you need to directly reply to: ``` $comment ```

Outputs

no outputs