agentsclimarketplace

Run3 Query GitHub Issues for December Activity

Skill cxcscmu/SkillLearnBench/skills/b3-teacher-feedback-claude-haiku-4-5/github-repo-analytics/run3_Query-GitHub-Issues-for-December-Activity

Fetch all issues created in December 2024 from cli/cli repository using correct date filtering, and identify bug-related issues based on label substring matching.From its SKILL.md

Install
npx -y skills add cxcscmu/SkillLearnBench --skill run3_Query-GitHub-Issues-for-December-Activity

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

SKILL.md

2.9 KB, 672 tokens by cl100k_base, as published. Nobody here has run it

Query GitHub Issues for December Activity

Overview

Use gh issue list to retrieve all issues created between 2024-12-01 and 2024-12-31 in the cli/cli repository, then filter for bug-related issues.

Step 1: Test the Issue Search Syntax

Verify the correct syntax for date range filtering with gh issue list:

gh issue list \
  --repo cli/cli \
  --state all \
  --search "created:2024-12-01..2024-12-31" \
  --json number,title,labels,createdAt,closedAt,state \
  --limit 200

Critical Notes:

  • gh issue list may NOT support --search flag the same way as gh api graphql
  • Test this command locally first to confirm it returns results
  • If results are roughly 2x expected, the date filter may be malfunctioning or including unintended issues
  • Verify --limit 200 is sufficient or increase as needed
  • Check that state=all or state=open,closed correctly filters the date range

Step 2: Verify Output Fields

Ensure the JSON output contains:

  • number: Issue identifier
  • labels: Array of label objects with name field (e.g., [{"name": "bug"}, {"name": "type: regression"}])
  • createdAt: ISO-8601 timestamp
  • closedAt: ISO-8601 timestamp or null (indicates whether issue is open)
  • state: Should indicate OPEN or CLOSED (verify actual values)

Step 3: Count Total Issues

Count the total number of issues returned by the query — this should be significantly lower than you may be getting (verify expected count ~30, not ~60).

Step 4: Identify Bug Reports

For each issue, check if ANY label contains the substring bug (case-insensitive or case-sensitive per repo convention):

  • Examples of matching labels: bug, type: bug, kind/bug, bug-report
  • Examples of non-matching labels: debugging, bugsquash (must contain as word/component, not substring if repo uses strict naming)
  • Count all issues with at least one matching label

Important: Verify the case sensitivity convention used in the cli/cli repo (test manually).

Step 5: Count Resolved Bug Reports

From the bug report issues identified in step 4:

  • Count only those where closedAt is NOT null OR state == "CLOSED"
  • These are bug reports that were closed during the month

Troubleshooting

  • If total count is ~60 instead of ~30: The date range filter is not working correctly; test the command locally
  • If bug count is ~28 instead of ~9: You may be including non-December issues or counting non-bug labels
  • If resolved_bugs is ~26 instead of ~8: Check if you're only counting bugs closed in December, or all closed bugs created in December
  • Consider switching to GraphQL if CLI filters prove unreliable: gh api graphql with explicit date range queries

What ships with it

Read from the repository

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

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.