一键导入
landscape-health
Batch health check all tracked projects — detect inactive repos, stale data, archived GitHub repos, and auto-suggest status changes.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Batch health check all tracked projects — detect inactive repos, stale data, archived GitHub repos, and auto-suggest status changes.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Add a new project to the AI Native Landscape from a GitHub URL or GitHub issue requesting inclusion.
Archive or deprecate a project in the AI Native Landscape. Set status to archived or inactive, record reason and date.
Bulk import multiple projects to the AI Native Landscape from a JSON file, CSV file, or list of GitHub URLs.
Generate statistics and analytics reports for the AI Native Landscape — category counts, license distribution, activity trends, and growth metrics.
Search the AI Native Landscape for open-source AI projects by natural language description. Find tools, frameworks, agents, runtimes, and infrastructure for AI native systems.
Full sync all landscape projects from GitHub — batch update metadata, detect status changes, and keep the entire catalog fresh.
| name | landscape-health |
| description | Batch health check all tracked projects — detect inactive repos, stale data, archived GitHub repos, and auto-suggest status changes. |
Scan all tracked projects for signs of abandonment, staleness, or issues. Generate a health report and suggest status changes.
landscape-sync to identify issuesgrep -rl "status: tracked" data/projects/*.en.md | sed 's/\.en\.md$//' | xargs -I{} basename {}
This gives all active slugs. Count them.
| Indicator | Healthy | Warning | Critical |
|---|---|---|---|
| Last commit | < 6 months | 6-12 months | > 12 months |
| GitHub archived | false | — | true |
| Open issues trend | Stable or declining | Slowly growing | Rapidly growing, no response |
| Last release | < 6 months | 6-18 months | > 18 months or never |
| Stars trend | Growing | Flat | Declining significantly |
For each tracked project, fetch activity data:
# For a single project
gh repo view <owner>/<repo> --json updatedAt,isArchived,stargazerCount,openIssues,latestRelease
# Batch: process all projects (respect rate limits)
Rate limit strategy:
gh api supports ~5000 requests/hour (authenticated)Collect for each project:
updatedAt — last push dateisArchived — GitHub archive statusstargazerCount — current starsopenIssues — open issue countlatestRelease.publishedAt — last release date (if any)For each project, calculate:
Activity score (0-100):
18 months: 0
Release cadence score (0-100):
12 months: 20
Overall health = (activity * 0.6 + release * 0.4)
| Classification | Criteria | Suggested action |
|---|---|---|
| Healthy | Score ≥ 60 | No action |
| Warning | Score 30-59 | Flag for review |
| Inactive | Score < 30 | Suggest status: inactive |
| Dead | GitHub archived or no commit > 18 months | Suggest status: archived |
Output a structured report:
Summary:
Landscape Health Report — YYYY-MM-DD
Total tracked: X
Healthy: X (X%)
Warning: X (X%)
Suggested inactive: X
Suggested archived: X
By category:
Category breakdown:
agents: X healthy / X warning / X inactive
rag-knowledge: X healthy / X warning / X inactive
...
Action items (sorted by urgency):
🔴 Archive immediately (GitHub archived):
- <slug> (repo archived on YYYY-MM-DD)
- <slug> (no activity since YYYY-MM-DD)
🟡 Review — suggest inactive:
- <slug> (last commit: YYYY-MM-DD, score: XX)
- <slug> (last commit: YYYY-MM-DD, score: XX)
🟢 Healthy (sample of highest scoring):
- <slug> (score: XX)
- <slug> (score: XX)
If user approves:
landscape-archive with reason no-activity-18m or repo-archivedstatus: inactive in both language filesAsk user: "Apply all suggested changes? [all/archive-only/review-each/none]"
Save the report to project root for reference:
# Save to reports/ directory
File: reports/health-YYYY-MM-DD.md
When user asks about one project's health:
landscape-health <slug>
Show detailed health breakdown for that project only.
For ongoing monitoring, this skill can be scheduled:
| Error | Fix |
|---|---|
| GitHub API rate limit | Process in batches, save progress, resume |
| Repo not found | Mark for archival, repo deleted |
| Network timeout | Retry with backoff, skip and flag |
gh not authenticated | Prompt user to run gh auth login |