Gh issue analysis

Skill cxcscmu/SkillLearnBench/skills/b4-skill-creator-claude-sonnet-4-6/github-repo-analytics/gh-issue-analysis

[COLM'26] SkillLearnBench is the first benchmark for evaluating continual learning methods that automatically generate agent skills.

Install

npx -y skills add cxcscmu/SkillLearnBench --skill gh-issue-analysis

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its author says it does

Copied from the file, not written here

How to fetch and analyze GitHub issues for a repository within a date range: total counts, bug report identification by label substring matching, and resolved bug counts. Use this skill whenever the user asks about issue metrics, bug counts, triage stats, or community health reports for a GitHub repo.

SKILL.md

3.0 KB, 661 tokens by cl100k_base, as published. Nobody here has run it

Issue Analysis via GitHub REST API

Fetching issues (excluding PRs)

The /repos/{owner}/{repo}/issues endpoint returns BOTH issues and PRs. Filter out PRs by checking for the absence of the pull_request key:

def get_issues(owner, repo, start_dt, end_dt, token=None):
    headers = {"Accept": "application/vnd.github+json"}
    if token:
        headers["Authorization"] = f"token {token}"

    results, page = [], 1
    while True:
        r = requests.get(f"https://api.github.com/repos/{owner}/{repo}/issues",
                         headers=headers,
                         params={"state": "all", "sort": "created",
                                 "direction": "desc", "per_page": 100,
                                 "page": page,
                                 "since": start_dt.isoformat()})
        data = r.json()
        if not data:
            break
        stop = False
        for item in data:
            if "pull_request" in item:
                continue   # skip PRs
            created = datetime.fromisoformat(item["created_at"].replace("Z", "+00:00"))
            if created > end_dt:
                continue
            if created < start_dt:
                stop = True
                break
            results.append(item)
        if stop or len(data) < 100:
            break
        page += 1
    return results

Identifying bug reports

A bug report is any issue where at least one label name contains the substring "bug" (case-insensitive):

def is_bug(issue):
    return any("bug" in label["name"].lower() for label in issue.get("labels", []))

bug_issues = [i for i in issues if is_bug(i)]

Common matching labels: bug, type: bug, kind/bug, confirmed-bug, etc.

Counting resolved bugs in the period

A bug is "resolved" (closed) during the period if:

is_bug(issue) is True
issue["state"] == "closed"
issue["closed_at"] falls within the target month

def parse_dt(s):
    return datetime.fromisoformat(s.replace("Z", "+00:00"))

resolved_bugs = [
    i for i in bug_issues
    if i["state"] == "closed"
    and i.get("closed_at")
    and start_dt <= parse_dt(i["closed_at"]) <= end_dt
]

Note: An issue created before December can be closed in December and still count as a resolved bug for the period. Fetch with state=all and filter closed_at in Python.

Gotchas

The since query param on /issues filters by updated_at, not created_at. Always re-filter in Python using created_at.
Labels may be empty ([]) for unlabelled issues — handle gracefully.
Rate limit: each page = 1 request. With 60 unauthenticated requests, you can fetch ~6000 items before hitting the limit.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.