Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
167 changes: 167 additions & 0 deletions .github/actions/claude-code-action/action.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,167 @@
name: "Claude Code Action"
description: "Run Claude Code in GitHub Actions workflows"

inputs:
github_token:
description: "GitHub token with repo and issues permissions"
required: true
anthropic_api_key:
description: "Anthropic API key"
required: true
cborg_api_key:
description: "CBORG API key"
required: true
prompt:
description: "The prompt to send to Claude Code"
required: false
default: ""
prompt_file:
description: "Path to a file containing the prompt to send to Claude Code"
required: false
default: ""
allowed_tools:
description: "Comma-separated list of allowed tools for Claude Code to use"
required: false
default: ""
output_file:
description: "File to save Claude Code output to (optional)"
required: false
default: ""
timeout_minutes:
description: "Timeout in minutes for Claude Code execution"
required: false
default: "10"
install_github_mcp:
description: "Whether to install the GitHub MCP server"
required: false
default: "false"
install_artl_mcp:
description: "Whether to install the ARTL MCP server"
required: false
default: "false"


runs:
using: "composite"
steps:
- name: Install uvx
shell: bash
run: |
curl -LsSf https://astral.sh/uv/install.sh | sh
source $HOME/.cargo/env
which uvx || echo "uvx not found in PATH, installing via pip"
pip install uv || true

- name: Install Claude Code
shell: bash
run: npm install -g @anthropic-ai/claude-code

- name: Install GitHub MCP Server
if: inputs.install_github_mcp == 'true'
shell: bash
run: |
claude mcp add-json github '{
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"GITHUB_PERSONAL_ACCESS_TOKEN",
"ghcr.io/github/github-mcp-server:sha-ff3036d"
],
"env": {
"GITHUB_PERSONAL_ACCESS_TOKEN": "${{ inputs.GITHUB_TOKEN }}"
}
}'

- name: Install ARTL MCP Server
if: inputs.install_artl_mcp == 'true'
shell: bash
run: |
claude mcp add-json artl '{
"command": "uvx",
"args": [
"artl-mcp"
]
}'

- name: Prepare Prompt File
shell: bash
id: prepare_prompt
run: |
# Check if either prompt or prompt_file is provided
if [ -z "${{ inputs.prompt }}" ] && [ -z "${{ inputs.prompt_file }}" ]; then
echo "::error::Neither 'prompt' nor 'prompt_file' was provided. At least one is required."
exit 1
fi

# Determine which prompt source to use
if [ ! -z "${{ inputs.prompt_file }}" ]; then
# Check if the prompt file exists
if [ ! -f "${{ inputs.prompt_file }}" ]; then
echo "::error::Prompt file '${{ inputs.prompt_file }}' does not exist."
exit 1
fi

# Use the provided prompt file
PROMPT_PATH="${{ inputs.prompt_file }}"
else
mkdir -p /tmp/claude-action
PROMPT_PATH="/tmp/claude-action/prompt.txt"
echo "${{ inputs.prompt }}" > "$PROMPT_PATH"
fi

# Verify the prompt file is not empty
if [ ! -s "$PROMPT_PATH" ]; then
echo "::error::Prompt is empty. Please provide a non-empty prompt."
exit 1
fi

# Save the prompt path for the next step
echo "PROMPT_PATH=$PROMPT_PATH" >> $GITHUB_ENV

- name: Run Claude Code
shell: bash
id: run_claude
run: |
ALLOWED_TOOLS_ARG=""
if [ ! -z "${{ inputs.allowed_tools }}" ]; then
ALLOWED_TOOLS_ARG="--allowedTools ${{ inputs.allowed_tools }}"
fi

# Set a timeout to ensure the command doesn't run indefinitely
timeout_seconds=$((${{ inputs.timeout_minutes }} * 60))

if [ -z "${{ inputs.output_file }}" ]; then
# Run Claude Code and output to console
timeout $timeout_seconds claude \
-p \
--verbose \
--output-format stream-json \
"$(cat ${{ env.PROMPT_PATH }})" \
${{ inputs.allowed_tools != '' && format('--allowedTools "{0}"', inputs.allowed_tools) || '' }}
else
# Run Claude Code and tee output to console and file
timeout $timeout_seconds claude \
-p \
--verbose \
--output-format stream-json \
"$(cat ${{ env.PROMPT_PATH }})" \
${{ inputs.allowed_tools != '' && format('--allowedTools "{0}"', inputs.allowed_tools) || '' }} | tee output.txt

# Process output.txt into JSON in a separate step
jq -s '.' output.txt > output.json

# Extract the result from the last item in the array (system message)
jq -r '.[-1].result' output.json > "${{ inputs.output_file }}"

echo "Complete output saved to output.json, final response saved to ${{ inputs.output_file }}"
fi

env:
ANTHROPIC_API_KEY: "."
Copy link

Copilot AI Sep 12, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The ANTHROPIC_API_KEY is set to a literal dot (".") which is likely incorrect. This should either use the anthropic_api_key input or be removed if CBORG authentication is the intended method.

Suggested change
ANTHROPIC_API_KEY: "."

Copilot uses AI. Check for mistakes.
ANTHROPIC_AUTH_TOKEN: ${{ inputs.cborg_api_key }}
GITHUB_TOKEN: ${{ inputs.github_token }}
ANTHROPIC_BASE_URL: "https://api.cborg.lbl.gov"
DISABLE_NON_ESSENTIAL_MODEL_CALLS: "1"
74 changes: 74 additions & 0 deletions .github/actions/claude-issue-summarize-action/action.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,74 @@
name: "Claude Issue summarize Action"
description: "Automatically Summarize GitHub issues using Claude Code"

inputs:
timeout_minutes:
description: "Timeout in minutes for execution"
required: false
default: "5"
anthropic_api_key:
description: "Anthropic API key"
required: true
cborg_api_key:
description: "CBORG API key"
required: true
github_token:
description: "GitHub token with repo and issues permissions"
required: true

runs:
using: "composite"
steps:
- name: Checkout repository code
uses: actions/checkout@v4
with:
fetch-depth: 0

- name: Create prompt file
shell: bash
run: |
mkdir -p /tmp/claude-prompts
cat > /tmp/claude-prompts/claude-issue-summarize-prompt.txt << 'EOF'
You're an Summarization assistant for GitHub issues. Your task is to analyze the issue, look up related issues, and summarize the issue.

Remember that this repo is primarily a note-taking and memory device for Chris Mungall, so anything you can do to enhance and enrich that would be helpful.

IMPORTANT: Don't make any PRs. You should ONLY post a response, IN THE SAME ISSUE.

Issue Information:
- REPO: ${{ github.repository }}
- ISSUE_NUMBER: ${{ github.event.issue.number }}

TASK OVERVIEW:

1. First, use your judgement to determine the course of action.
- If the issue is an epic or has links to other issues, read those issues, and summarize them. Summarize their status etc. Given an overall summary.
- If the issue is an idea or a vague high level description, run with the idea, do deep research, give heloful suggestions with links to relevant papers etc.
- If there are links in the issue to websites, database entities, etc, then look these up.
- If there are links or references to papers, then looks these up and summarize (don't guess content - if you can't access the paper, then say "I can't access the paper")
- Try using artl-mcp to get paper fulltext.
- If you want a more powerful AI agent to take on this one and make a PR, use this magic phrase: `@dragon-ai-agent please <INSTRUCTIONS>`

2. Next, use the GitHub tools to get context about the issue:
- You have access to these github tools:
- mcp__github__get_issue: Use this to retrieve the current issue's details including title, description, and existing labels
- mcp__github__get_issue_comments: Use this to read any discussion or additional context provided in the comments
- mcp__github__update_issue: Use this to apply labels to the issue (do not use this for commenting)
- mcp__github__search_issues: Use this to find similar issues that might provide context for proper categorization and to identify potential duplicate issues
- mcp__github__list_issues: Use this to understand patterns in how other issues are labeled
- You can also use web searching and fetching.

- It's okay to not add any information if the issue is not clear.
EOF

- name: Run Claude Code
uses: ./.github/actions/claude-code-action
with:
prompt_file: /tmp/claude-prompts/claude-issue-summarize-prompt.txt
allowed_tools: "Bash(gh label list),WebFetch,Fetch,LS,mcp__github__get_issue,mcp__github__get_issue_comments,mcp__github__update_issue,mcp__github__add_issue_comment,mcp__github__search_issues,mcp__github__list_issues"
install_github_mcp: "true"
install_artl_mcp: "true"
timeout_minutes: ${{ inputs.timeout_minutes }}
anthropic_api_key: ${{ inputs.anthropic_api_key }}
cborg_api_key: ${{ inputs.cborg_api_key }}
github_token: ${{ inputs.github_token }}
91 changes: 91 additions & 0 deletions .github/actions/claude-issue-triage-action/action.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,91 @@
name: "Claude Issue Triage Action"
description: "Automatically triage GitHub issues using Claude Code"

inputs:
timeout_minutes:
description: "Timeout in minutes for execution"
required: false
default: "5"
anthropic_api_key:
description: "Anthropic API key"
required: true
cborg_api_key:
description: "CBORG API key"
required: true
github_token:
description: "GitHub token with repo and issues permissions"
required: true

runs:
using: "composite"
steps:
- name: Checkout repository code
uses: actions/checkout@v4
with:
fetch-depth: 0

- name: Create prompt file
shell: bash
run: |
mkdir -p /tmp/claude-prompts
cat > /tmp/claude-prompts/claude-issue-triage-prompt.txt << 'EOF'
You're an issue triage assistant for GitHub issues. Your task is to analyze the issue and select appropriate labels from the provided list.

IMPORTANT: Don't post any comments or messages to the issue. Your only action should be to apply labels.

Issue Information:
- REPO: ${{ github.repository }}
- ISSUE_NUMBER: ${{ github.event.issue.number }}

TASK OVERVIEW:

1. First, fetch the list of labels available in this repository by running: `gh label list`. Run exactly this command with nothing else.

2. Next, use the GitHub tools to get context about the issue:
- You have access to these tools:
- mcp__github__get_issue: Use this to retrieve the current issue's details including title, description, and existing labels
- mcp__github__get_issue_comments: Use this to read any discussion or additional context provided in the comments
- mcp__github__update_issue: Use this to apply labels to the issue (do not use this for commenting)
- mcp__github__search_issues: Use this to find similar issues that might provide context for proper categorization and to identify potential duplicate issues
- mcp__github__list_issues: Use this to understand patterns in how other issues are labeled
- Start by using mcp__github__get_issue to get the issue details

3. Analyze the issue content, considering:
- The issue title and description
- The type of issue (bug report, feature request, question, etc.)
- Technical areas mentioned
- Severity or priority indicators
- User impact
- Components affected

4. Select appropriate labels from the available labels list provided above:
- Choose labels that accurately reflect the issue's nature
- Be specific but comprehensive
- Select priority labels if you can determine urgency (high-priority, med-priority, or low-priority)
- Consider platform labels (android, ios) if applicable
- If you find similar issues using mcp__github__search_issues, consider using a "duplicate" label if appropriate. Only do so if the issue is a duplicate of another OPEN issue.

5. Apply the selected labels:
- Use mcp__github__update_issue to apply your selected labels
- DO NOT post any comments explaining your decision
- DO NOT communicate directly with users
- If no labels are clearly applicable, do not apply any labels

IMPORTANT GUIDELINES:
- Be thorough in your analysis
- Only select labels from the provided list above
- DO NOT post any comments to the issue
- Your ONLY action should be to apply labels using mcp__github__update_issue
- It's okay to not add any labels if none are clearly applicable
EOF

- name: Run Claude Code
uses: ./.github/actions/claude-code-action
with:
prompt_file: /tmp/claude-prompts/claude-issue-triage-prompt.txt
allowed_tools: "Bash(gh label list),mcp__github__get_issue,mcp__github__get_issue_comments,mcp__github__update_issue,mcp__github__search_issues,mcp__github__list_issues"
install_github_mcp: "true"
timeout_minutes: ${{ inputs.timeout_minutes }}
anthropic_api_key: ${{ inputs.anthropic_api_key }}
cborg_api_key: ${{ inputs.cborg_api_key }}
github_token: ${{ inputs.github_token }}
1 change: 1 addition & 0 deletions .github/ai-controllers.json
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
["realmarcin"]
1 change: 1 addition & 0 deletions .github/copilot-instructions.md
40 changes: 40 additions & 0 deletions .github/copilot-setup-steps.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,40 @@
name: "Copilot Setup Steps"


# Automatically run the setup steps when they are changed to allow for easy validation, and
# allow manual testing through the repository's "Actions" tab
on:
workflow_dispatch:
push:
paths:
- .github/workflows/copilot-setup-steps.yml
pull_request:
paths:
- .github/workflows/copilot-setup-steps.yml

jobs:
# The job MUST be called `copilot-setup-steps` or it will not be picked up by Copilot.
copilot-setup-steps:
runs-on: ubuntu-latest

# Set the permissions to the lowest permissions possible needed for your steps.
# Copilot will be given its own token for its operations.
permissions:
# If you want to clone the repository as part of your setup steps, for example to install dependencies, you'll need the `contents: read` permission. If you don't clone the repository in your setup steps, Copilot will do this for you automatically after the steps complete.
contents: read

# You can define any steps you want, and they will run before the agent starts.
# If you do not check out your code, Copilot will do this for you.
steps:
- name: Checkout code
uses: actions/checkout@v4

- name: Install uv
uses: astral-sh/setup-uv@v5

- name: Install Python tools
run: |
uv sync
source .venv/bin/activate
shell: bash

Loading
Loading