-
Notifications
You must be signed in to change notification settings - Fork 0
feat: workflow の sdk 記述方法を変更 #9
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Merged
Merged
Changes from all commits
Commits
Show all changes
2 commits
Select commit
Hold shift + click to select a range
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -14,7 +14,7 @@ | |
|
|
||
| import boto3 | ||
| from botocore.config import Config | ||
| from claude_agent_sdk import ClaudeAgentOptions, query, AssistantMessage, TextBlock, ToolUseBlock | ||
| from claude_agent_sdk import ClaudeAgentOptions, ClaudeSDKClient, query, AssistantMessage, TextBlock, ToolUseBlock | ||
| from claude_agent_sdk.types import StreamEvent | ||
|
|
||
|
|
||
|
|
@@ -99,36 +99,18 @@ def s3_upload_text(s3, bucket, key, text, content_type="text/plain"): | |
| raise | ||
|
|
||
|
|
||
| async def implement_proposal( | ||
| repo_dir: str, proposal_plan: str, screenshot_output: str | ||
| ) -> None: | ||
| """Use Claude Agent SDK to implement the design proposal, start dev server, and take screenshot.""" | ||
| async def implement_changes(repo_dir: str, proposal_plan: str) -> None: | ||
| """Use Claude Agent SDK to implement the design proposal.""" | ||
| prompt = f"""You are implementing UI changes to a web application located at {repo_dir}. | ||
|
|
||
| Here is the specific design proposal to implement: | ||
|
|
||
| {proposal_plan} | ||
|
|
||
| Follow these steps in order: | ||
|
|
||
| ## Step 1: Implement the changes | ||
| - Implement all the changes described in the plan | ||
| - All file modifications must be correct and complete | ||
| - Follow existing code conventions and patterns | ||
| - Do not leave any TODO comments or incomplete implementations | ||
|
|
||
| ## Step 2: Start the dev server and take a screenshot | ||
| After implementing the changes: | ||
| 1. Investigate how to start the dev server for this project (check package.json scripts, README, etc.) | ||
| 2. Install dependencies if needed | ||
| 3. Start the dev server in the background using Bash (e.g. `npm run dev &` or whatever is appropriate) | ||
| 4. Wait for the server to become ready (poll with curl until it responds) | ||
| 5. Once the server is ready, take a screenshot by running: | ||
| `node /usr/local/bin/take-screenshot.mjs <server_url> {screenshot_output}` | ||
| where <server_url> is the URL the dev server is listening on (e.g. http://localhost:5173) | ||
| 6. Make sure the screenshot file was created at {screenshot_output} | ||
|
|
||
| IMPORTANT: The screenshot is critical. Do your best to get the dev server running and take the screenshot. | ||
| """ | ||
|
|
||
| current_tool = None | ||
|
|
@@ -163,7 +145,6 @@ async def implement_proposal( | |
| tool_input_chunks = "" | ||
| continue | ||
|
|
||
| # Log progress messages from completed AssistantMessage | ||
| if isinstance(msg, AssistantMessage): | ||
| for block in msg.content: | ||
| if isinstance(block, TextBlock): | ||
|
|
@@ -172,6 +153,82 @@ async def implement_proposal( | |
| _emit_tool_detail("implementing", block.name, json.dumps(block.input)) | ||
|
|
||
|
|
||
| async def _process_messages(client: ClaudeSDKClient, phase: str) -> None: | ||
| """Process messages from ClaudeSDKClient, emitting logs for a given phase.""" | ||
| current_tool = None | ||
| tool_input_chunks = "" | ||
|
|
||
|
Comment on lines
+156
to
+160
|
||
| async for msg in client.receive_response(): | ||
| if isinstance(msg, StreamEvent): | ||
| event = msg.event | ||
| etype = event.get("type") | ||
| if etype == "content_block_start": | ||
| content_block = event.get("content_block", {}) | ||
| if content_block.get("type") == "tool_use": | ||
| current_tool = content_block.get("name") | ||
| tool_input_chunks = "" | ||
| elif etype == "content_block_delta": | ||
| delta = event.get("delta", {}) | ||
| if delta.get("type") == "input_json_delta": | ||
| tool_input_chunks += delta.get("partial_json", "") | ||
| elif etype == "content_block_stop": | ||
| if current_tool and tool_input_chunks: | ||
| _emit_tool_detail(phase, current_tool, tool_input_chunks) | ||
| current_tool = None | ||
| tool_input_chunks = "" | ||
| continue | ||
|
|
||
| if isinstance(msg, AssistantMessage): | ||
| for block in msg.content: | ||
| if isinstance(block, TextBlock): | ||
| emit_log(phase, block.text[:200], detail=block.text) | ||
| elif isinstance(block, ToolUseBlock): | ||
| _emit_tool_detail(phase, block.name, json.dumps(block.input)) | ||
|
|
||
|
|
||
| async def launch_and_screenshot(repo_dir: str, screenshot_output: str) -> None: | ||
| """Launch dev server and take screenshot in a single session. | ||
|
|
||
| Uses ClaudeSDKClient to keep the same conversation context across | ||
| two query() calls so the agent remembers the dev server URL. | ||
| """ | ||
| options = ClaudeAgentOptions( | ||
| allowed_tools=["Read", "Bash", "Glob", "Grep"], | ||
| cwd=repo_dir, | ||
| max_turns=20, | ||
| max_budget_usd=1.0, | ||
| include_partial_messages=True, | ||
| ) | ||
|
|
||
| async with ClaudeSDKClient(options=options) as client: | ||
| # Phase 1: install deps + start dev server | ||
| emit_log("launching", "Launching project") | ||
| await client.query(f"""You need to launch the web application at {repo_dir}. | ||
|
|
||
| Follow these steps: | ||
| 1. Investigate how to start the dev server (check package.json scripts, README, etc.) | ||
| 2. Install the project's dependencies if needed (e.g. `npm install` in the project directory) | ||
| 3. Start the dev server in the background using Bash (e.g. `npm run dev &` or whatever is appropriate) | ||
| 4. Wait for the server to become ready (poll with curl until it responds) | ||
|
|
||
| IMPORTANT: Make sure the dev server is running and responding before finishing. | ||
| """) | ||
| await _process_messages(client, "launching") | ||
|
|
||
| # Phase 2: take screenshot (same session, so dev server URL is remembered) | ||
| emit_log("screenshot", "Taking after screenshot") | ||
| await client.query(f"""Now take a screenshot of the running application. | ||
|
|
||
| You already know the dev server URL from the previous step. Use it to take a screenshot: | ||
| 1. Run: `node /usr/local/bin/take-screenshot.mjs <server_url> {screenshot_output}` | ||
| where <server_url> is the URL the dev server is listening on | ||
| 2. Verify the screenshot file was created at {screenshot_output} | ||
|
|
||
| NOTE: Playwright and Chromium are pre-installed globally in this container. Do NOT install playwright or run `npx playwright install`. Just run `node /usr/local/bin/take-screenshot.mjs <url> <output>` directly. | ||
| """) | ||
| await _process_messages(client, "screenshot") | ||
|
|
||
|
|
||
| async def main() -> None: | ||
| # Read environment variables | ||
| session_id = os.environ["SESSION_ID"] | ||
|
|
@@ -261,7 +318,10 @@ async def main() -> None: | |
|
|
||
| screenshot_path = f"{tmp_dir}/after.png" | ||
| emit_log("implementing", "Implementing design proposal") | ||
| await implement_proposal(repo_dir, proposal_plan, screenshot_path) | ||
| await implement_changes(repo_dir, proposal_plan) | ||
|
|
||
| # Launch + screenshot in a single session (shares dev server URL context) | ||
| await launch_and_screenshot(repo_dir, screenshot_path) | ||
|
|
||
| # Step 4: Generate cumulative patch (squash everything since base_branch) | ||
| emit_log("uploading", "Generating cumulative patch") | ||
|
|
||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
_process_messages()(andlaunch_and_screenshot()below) duplicates the same Claude streaming/tool-log parsing code that now also exists indocker/worker-implement.py. This duplication increases maintenance cost and the chance that a future SDK change (event types / fields) is applied in only one worker. Consider moving the shared logic into a common helper module and importing it from both scripts.