🏥 Organization Health Report — github — 2026-03-23

[github-actions[bot]]

The github organization currently has 100 public, non-archived repositories tracked in this analysis. Overall, there are 6,444 open issues + PRs across all public repos. Recent activity (last 30 days) shows strong engineering momentum especially in the gh-aw, gh-aw-mcpg, and gh-aw-firewall repositories. The top 3 repositories by open backlog — codeql (1,323), copilot-cli (1,239), and spec-kit (625) — represent the highest-attention areas for maintainers.

📊 Volume Metrics

Overall Status (from repository metadata, 100 public repos)

Metric	Count
Public Non-Archived Repositories	100
Total Open Issues + PRs (all repos)	6,444
Sample Issues Analyzed (30d window)	87
Sample PRs Analyzed (30d window)	88

Recent Activity in Sample (Last 7 Days)

Metric	Count
Issues Created	86
Issues Closed	36
PRs Created	82
PRs Closed	47

Recent Activity in Sample (Last 30 Days)

Metric	Count
Issues Created	86
Issues Closed	36
Open Issues (in sample)	51
Open PRs (in sample)	41
Closed PRs	47

---

The charts above show issues and PR activity for the top repositories and authors over the last 30 days.

---

🏆 Top 10 Most Active Repositories

Ranked by weighted activity score (Issues × 2 + PRs × 3 + Comments × 0.5):

Repository	Issues	PRs	Comments	Score
gh-aw	44	40	506	461
gh-aw-mcpg	38	27	56	185
gh-aw-firewall	4	1	240	131
dependabot-action	0	3	0	9
copilot-cli	1	0	13	8
opensource.guide	0	2	3	8
awesome-copilot	0	2	1	7
markdownlint-github	0	2	0	6
eslint-plugin-github	0	2	0	6
codeql-coding-standards	0	2	0	6

Note: gh-aw, gh-aw-mcpg, and gh-aw-firewall dominate activity due to agentic workflow automation generating issues and tracking runs.

Top 6 Most Active Authors

Author	Issues Opened	PRs Opened	Score	Notes
github-actions[bot]	86	34	274	Automation
Copilot	0	29	87	AI agent
dependabot[bot]	0	16	48	Dependency updates
lpcox	0	7	21	Human contributor
jeongsoolee09	0	2	6	Human contributor
RyanHecht	1	0	2	Human contributor

The majority of activity is driven by automation (bots and AI agents). Human contributors are active primarily through PRs.

---

🔥 High-Activity Items Needing Attention

Hot Issues (Open, High Discussion)

Issue	Repository	Comments	Age	Link
[aw] No-Op Runs	gh-aw	211	5d	#21483
[agentics] No-Op Runs	gh-aw-firewall	189	3d	#1377
[agentics] Secret Digger (Codex) failed	gh-aw-firewall	27	3d	#1379
[aw] Issue Monster failed	gh-aw	25	3d	#21864
[agentics] Secret Digger (Copilot) failed	gh-aw-firewall	18	3d	#1378
[safeoutputs] Improve update_issue tool	gh-aw	16	0d	#22282
[performance] Regression in CompileMemoryUsage: +103.7%	gh-aw	16	0d	#22277
[aw] No-Op Runs	gh-aw-mcpg	16	3d	#2224

Hot PRs (Open, Active Discussion)

PR	Repository	Comments	Age	Link
Add builtin qmd documentation search tool	gh-aw	86	1d	#22183
feat: add label update support to safe-output	gh-aw	6	0d	#22384
chore(deps): bump h3 from 1.15.6 to 1.15.9	gh-aw-firewall	4	2d	#1390
chore(deps): bump nokogiri (opensource.guide)	opensource.guide	3	34d	#3610

⚠️ PR #22183 (86 comments) is highly active and likely needs a final review decision.

---

📉 Org-Wide Backlog by Repository

Repositories color-coded by backlog severity. Red = high backlog (>500), Orange = medium (200–500), Blue = healthy (<200).

Top Repositories by Open Backlog

Repository	Open Issues+PRs	Stars	Status
codeql	1,323	9,373	🔴 High
copilot-cli	1,239	9,514	🔴 High
spec-kit	625	81,054	🔴 High
gh-ost	323	13,259	🟠 Medium
github-mcp-server	315	28,168	🟠 Medium
docs	182	19,164	🔵 Healthy
copilot-sdk	154	7,972	🔵 Healthy
gh-aw	165	4,147	🔵 Healthy

---

⚠️ Items Needing Attention

Category	Count	Notes
Unassigned open issues (sample)	50/51	Nearly all open issues lack an assignee
Stale issues (30d+ old, inactive 7d+)	0	None in current sample
Stale PRs (30d+ old, inactive 7d+)	0	None in current sample
Unlabeled issues	0	All sampled issues have labels

The high unassigned count is expected given most issues are automation-generated tracking issues.

---

💡 Commentary and Recommendations

1. Agentic Automation is the Dominant Activity Driver
The majority of issues and PRs are created by github-actions[bot], Copilot AI agent, and dependabot[bot]. This is expected for repos running agentic workflows. Human contributors (e.g., lpcox, jeongsoolee09, RyanHecht) primarily contribute via PRs.

2. Three Repos Need Backlog Attention
codeql (1,323), copilot-cli (1,239), and spec-kit (625) have substantial open issue backlogs relative to their size. Maintainers should consider:

• Triaging and closing stale issues with labels like wontfix or needs-more-info
• Running batch-close for issues that are duplicates or outdated

3. PR #22183 Needs Resolution
The qmd documentation search tool PR in gh-aw has accumulated 86 comments in 1 day — it needs a clear merge/close decision to unblock the author and reviewers.

4. Performance Regression in gh-aw
Issue #22277 reports a +103.7% regression in CompileMemoryUsage — this should be prioritized before the next release.

5. No-Op Tracking Issues are High-Volume
The top two issues by comment count (gh-aw#21483 with 211 comments and gh-aw-firewall#1377 with 189 comments) are no-op tracking issues. Consider archiving or rotating these periodically to keep them manageable.

---

Data Collection Methodology

Data Collection

• Report Date: 2026-03-23
• Analysis Period: Last 30 days (2026-02-21 to 2026-03-23)
• Repositories Discovered: 100 public, non-archived repositories (first page from GitHub search)
• Issues Analyzed: 87 unique issues (org-wide search, last 30 days)
• PRs Analyzed: 88 unique PRs (org-wide search, last 30 days)
• Backlog Data: From GitHub repository metadata (open_issues_count field)

Methodology

• Repository discovery via GitHub MCP search_repositories with org:github archived:false
• Issue/PR data via GitHub MCP search_issues and search_pull_requests with org-wide queries
• Activity scores: Issues × 2 + PRs × 3 + Comments × 0.5
• Charts generated with Python + Matplotlib + Seaborn (DPI 150)
• Aggregate open issue counts sourced from repository metadata (includes both issues and PRs)

Limitations

• The 100-repository limit captures one page of results; the org has more public repos
• Issue/PR samples are bounded by API pagination (100 per query)
• Stale detection based on sampled data only — may not reflect full org picture

References:

• §23428187836 — Workflow run that generated this report
• github org public repos

AI generated by Organization Health Report · history

• expires on Mar 24, 2026, 8:38 AM UTC