Helper functions to add all issues of a repository to a GitHub project (beta), with example usage at bottom of script. For https://github.com/github/feedback/discussions/6765

populate-github-project-beta.py

import requests
import json


GH_GRAPHQL_URL = 'https://api.github.com/graphql'
TOKEN = '' # add your PAT with read:org and write:org scope

def get_json_result(query):
    headers={'Authorization': f'token {TOKEN}'}
    r = requests.post(GH_GRAPHQL_URL, json={'query': query}, headers=headers)
    if r.status_code != 200:
        raise Exception()
    return json.loads(r.text)

def add_issue_to_project(issue_id, project_id):
    mutation = f"""mutation {{
        addProjectNextItem(input: {{projectId: "{project_id}" contentId: "{issue_id}"}}) {{
            projectNextItem {{
                id
            }}
        }}
    }}
    """
    return get_json_result(mutation)

def get_project_id(org_name, project_number):
    query = f"""query {{
        organization(login: "{org_name}") {{
            projectNext(number: {project_number}) {{
                id
            }}
        }}
    }}
    """
    result = get_json_result(query)
    return result['data']['organization']['projectNext']['id']

def get_issue_ids(repo_name, owner_name, n=20):
    query = f"""query {{
        repository(owner: "{owner_name}", name: "{repo_name}") {{
            openIssues: issues(first: {n}) {{
                edges {{
                    node {{
                        number
                        id
                    }}
                }}
            }}
        }}
    }}
    """
    result = get_json_result(query)
    return [item['node']['id'] for item in result['data']['repository']['openIssues']['edges']]

# example: add 30 issues from serratus-bio/serratus.io to serratus-bio/projects/2.
project_id = get_project_id(org_name="serratus-bio", project_number=2)
issue_ids = get_issue_ids(repo_name="serratus.io", owner_name="serratus-bio", n=30)
for issue_id in issue_ids:
    add_issue_to_project(issue_id, project_id)

Thanks for this! The code needs changing if using a private personal project rather than an organization project.

• get_project_id should read user instead of organization on line 28.
• the personal access token also needs the repo permission (access private repos; apparently needed)

The maximum value for n on line 57 is 100 (can't get more than a hundred issues at once). Weirdly, even though the code says openIssues, it will get closed issues too... Anyway, this only gets the first "page" of results; the other issues after the first n are ignored, and the code doesn't allow fetching them. I don't know enough about the GraphQL API to know how to do this.

Ok, to get pagination working, you need to add the cursor field to the query for issues, then read the value of cursor for the last returned issue, and set this to the after parameter of the issues query.

def get_issue_ids(repo_name, owner_name, n=100, next_cursor=None):
    after_query = ''
    if next_cursor is not None:
        after_query = f'after: "{next_cursor}"'

    query = f"""query {{
        repository(owner: "{owner_name}", name: "{repo_name}") {{
            openIssues: issues(first: {n} {after_query}) {{
                edges {{
                    node {{
                        number
                        id
                    }}
                    cursor
                }}
            }}
        }}
    }}
    """

    result = get_json_result(query)
    edges = result['data']['repository']['openIssues']['edges']
    if len(edges) == 0:
        return [], None

    return [item['node']['id'] for item in edges], edges[-1]['cursor']

def get_all_issue_ids(repo_name, owner_name):
    issues = []
    next_cursor = None
    while True:
        new_issues, next_cursor = get_issue_ids(repo_name, owner_name, next_cursor=next_cursor)
        if len(new_issues) == 0:
            break

        issues.append(new_issues)

    return issues