GitHub

API

The GitHub connector is an API connector that imports issues and pull requests from a GitHub repository. It produces structural claims and optionally extracts additional claims from issue and PR bodies via text extraction.

How it works

Uses the GitHub REST API v3 with Bearer token authentication
Fetches issues and pull requests ordered by updated descending
Pagination handled automatically with per_page=100
Structural claims are produced for each issue/PR with confidence 1.0
Optionally extracts claims from issue/PR bodies via heuristic, smart, llm, or none extraction modes

Claim patterns

Subject	Predicate	Object
`(repo#number, issue)`	`authored_by`	`(author, person)`
`(repo#number, issue)`	`has_state`	`(open/closed, status)`
`(repo#number, issue)`	`labeled`	`(label, label)` per label
`(repo#number, issue)`	`assigned_to`	`(assignee, person)` per assignee

All structural claims have confidence 1.0. Claims extracted from bodies inherit the confidence of the extraction mode.

1 Create a Personal Access Token

Go to github.com/settings/tokens and generate a new token (classic or fine-grained).

2 Select scopes

repo — required for private repositories
public_repo — sufficient for public repositories only

3 Install the requests package

$ pip install requests

4 Connect

import os
conn = db.connect("github",
    token=os.environ["GITHUB_TOKEN"],
    repo="owner/repo",
)
result = conn.run(db)

Parameter	Required	Default	Description
`token`	Yes	—	GitHub personal access token
`repo`	Yes	—	Repository in `owner/repo` format
`include_prs`	No	`True`	Include pull requests
`state`	No	`all`	`all`, `open`, or `closed`
`max_items`	No	`500`	Maximum issues/PRs to fetch
`labels`	No	`None`	Filter to issues with these labels
`extraction`	No	`heuristic`	Text extraction mode for bodies: `heuristic`, `smart`, `llm`, or `none`
`save`	No	`False`	Encrypt and persist token

Basic usage

conn = db.connect("github",
    token=os.environ["GITHUB_TOKEN"],
    repo="owner/repo",
)
result = conn.run(db)

Issues only, filtered

conn = db.connect("github",
    token=os.environ["GITHUB_TOKEN"],
    repo="owner/repo",
    include_prs=False,
    state="open",
    labels=["bug", "priority"],
    max_items=100,
)
result = conn.run(db)

With LLM extraction and save

conn = db.connect("github",
    token=os.environ["GITHUB_TOKEN"],
    repo="owner/repo",
    extraction="llm",
    save=True,
)
result = conn.run(db)