Article Relevance Audit — Week of 2026-04-03

[github-actions]

Weekly health check across all published evergreen articles. 63 articles scanned (2 new since last run). Combined report covering all 4 audit categories.

---

🔗 Link Health Report

Articles scanned: 63 | New issues found: 5 | Persistent issues: 3

🆕 New — Unreachable Outbound Links

Article	Link Text	URL	Status	Notes
ci-monitor-extension-agent-ci-feedback-loop	prompts/ci-monitor-extension.md	(htek.dev/redacted)	Unreachable	/prompts/ path doesn't appear to exist on htek.dev — 404 or not deployed
ci-monitor-extension-agent-ci-feedback-loop	OpenShell sandboxing docs	(docs.openclaw.ai/redacted)	Unreachable	OpenClaw docs site down or not yet deployed
ci-monitor-extension-agent-ci-feedback-loop	gateway/sandboxing#openshell-backend	(docs.openclaw.ai/redacted)	Unreachable	Same — openclaw docs unreachable
who-needs-openclaw-copilot-cli-extensions	OpenShell sandboxing	(docs.openclaw.ai/redacted)	Unreachable	Same openclaw docs issue
who-needs-openclaw-copilot-cli-extensions	OpenClaw root	(docs.openclaw.ai/redacted)	Unreachable	Same

⚠️ Broken Internal Link — PERSISTENT (4th Audit, Unfixed Since 2026-03-11)

Article	Link Text	Broken URL	Fix
agentic-ai-transforming-dev-teams	"GitHub Copilot CLI worktrees"	/articles/copilot-fleet-mode-video-pipeline	Change to /articles/video-pipeline-with-fleet-mode

This link has been flagged in every audit since March 11. The correct slug (video-pipeline-with-fleet-mode) exists. This is a one-line fix.

🆕 Draft Articles Being Linked

Article	Linked Slug	Issue
agentic-ai-transforming-dev-teams	copilot-cli-native-sandboxing-agent-security	Article is draft: true — readers get a 404
microsoft-ai-agent-security-first-class-infrastructure	copilot-cli-native-sandboxing-agent-security	Same — new link found this audit

Persistent — Unreachable Outbound (3rd+ Audit)

Article	URL	Status
git-worktree-unlocks-agentic-development	(resources.anthropic.com/redacted)	Unreachable — PDF may be gated or URL incorrect
agent-harnesses-controlling-ai-agents-2026	(www.anthropic.com/redacted)	Unreachable — page may have moved or been removed

---

📊 Factual Freshness Report

Articles scanned: 63 | Stale references found: 6 (5 persistent, 1 new)

Potentially Stale References

Article	What It Says	Current Reality	Severity	Status
ai-fixes-its-own-bugs	"SWE-agent solves 12.47% of SWE-bench"	Current SOTA is significantly higher; SWE-bench leaderboard has moved substantially since 2025	🔴 High	Persistent — 4th audit unfixed
ai-fixes-its-own-bugs	"CodeQL Autofix powered by GPT-4o"	GPT-4o is two generations old (GPT-5.x is current); CodeQL Autofix model may have been updated	🟡 Medium	Persistent
copilot-developer-fulfillment	"20 million GitHub Copilot users"	GitHub likely announced updated user counts at GitHub Universe or in recent earnings	🟡 Medium	Persistent
copilot-developer-fulfillment	"2024 DORA report"	2025 DORA report likely published	🟡 Medium	Persistent
choosing-the-right-ai-sdk	"GitHub Copilot SDK: Technical Preview (Jan 2026)"	3 months have passed — SDK may have reached GA or updated significantly	🟡 Medium	Persistent
az-prototype-scaffold-enterprise-azure-with-copilot	"Stable release is still tracking for next week" (written ~Mar 27)	It's now April 3 — the predicted window has passed; no evidence of stable release yet	🟡 Medium	🆕 New this run

---

🔀 Cross-Linking Opportunities

New articles with zero inbound links: 2 | Actionable opportunities: 3

ci-monitor-extension-agent-ci-feedback-loop

File: src/content/articles/ci-monitor-extension-agent-ci-feedback-loop.mdx
Published: 2026-03-31 | Inbound links: 0

1. Link from: copilot-cli-extensions-cookbook-examples (2026-03-26)
   • Suggested section: ### Example 16: The REPL Loop — Self-Healing Until Green
   • Insert near: "You can swap npm test for any validation: npm run lint, npm run build... I've used this pattern to get the agent to write a feature, run the tests, fix failures, re-run, and ship — all from a single prompt."
   • Suggested text: I took this pattern further in production — [I built a full CI monitor extension](/articles/ci-monitor-extension-agent-ci-feedback-loop) that wraps GitHub Actions into the loop, so the agent watches CI, reads failure logs, and iterates without me babysitting the terminal.

who-needs-openclaw-copilot-cli-extensions

File: src/content/articles/who-needs-openclaw-copilot-cli-extensions.mdx
Published: 2026-04-03 | Inbound links: 0

2. Link from: nvidia-openshell-sandbox-ai-agents (2026-03-23)

   • Suggested section: Near the closing section on real-world applications
   • Insert near: "That's the part that changes everything: policy as code."
   • Suggested text: The real-world payoff of that security infrastructure is starting to show — [I built a Telegram bridge to Copilot CLI in ~420 lines](/articles/who-needs-openclaw-copilot-cli-extensions) that runs inside an OpenShell sandbox, using the L7 proxy for credential injection. Same agent, now remote-controlled from your phone.

3. Link from: copilot-cli-remote-access-your-agent-from-anywhere (2026-03-23)

   • Suggested section: Near session continuity or remote access patterns
   • Insert near: Any passage discussing accessing the agent from other devices
   • Suggested text: I took this further by [building a Telegram bridge in a single extension file](/articles/who-needs-openclaw-copilot-cli-extensions) — no gateway, no daemon, just the extension SDK and a bot token. Full Copilot CLI, remote-controlled from your phone.

Persistent Zero-Inbound Articles (Carryover)

These articles have had zero inbound links for multiple audit cycles:
github-copilot-vs-the-world, stop-pulling-your-agentic-unicorns-off-the-field, vibe-testing-when-ai-agents-goodhart-your-test-suite, specs-equal-tests-terraform-ai-development, cryptographic-approval-gates-ai-agents, ai-audit-codebase-testability, github-copilot-visual-studio-guide, lessons-from-500-github-migrations, vidpipe-copilot-cli-challenge

---

🌿 Evergreen Maintenance

Refresh candidates: 5

Refresh Candidates

Article	Published	Last Updated	Topic Area	Why Refresh?	Priority
agentic-ai-transforming-dev-teams	2026-02-20	Never	AI dev teams	Has a broken link (4th audit); new convergent architecture patterns, Stripe/Coinbase case studies published	🔴 High
ai-fixes-its-own-bugs	2026-02-14	Never	AI code repair	SWE-bench leaderboard moved significantly; CodeQL Autofix model outdated	🔴 High
az-prototype-scaffold-enterprise-azure-with-copilot	2026-03-23	Never	Azure prototyping	"Stable release next week" prediction has passed — needs status update	🟡 Medium
context-engineering-key-to-ai-development	2026-02-14	Never	Context engineering	MCP has grown substantially; Karpathy's framing, new Copilot CLI extensions SDK all relevant	🟡 Medium
choosing-the-right-ai-sdk	2026-02-14	Never	AI SDKs	GitHub Copilot SDK Technical Preview → possible GA; new SDKs launched (Anthropic Claude Code SDK, etc.)	🟡 Medium

Detailed Refresh Suggestion

az-prototype-scaffold-enterprise-azure-with-copilot

The article explicitly says "Stable release is still tracking for next week" — written around March 27. It's now April 3. The prediction window has passed. This needs either:

• A follow-up update section confirming GA shipped, OR
• An update noting the release was delayed beyond the predicted window

Either way, leaving a time-bound prediction unaddressed erodes credibility. This is a quick fix — one paragraph update in the existing "Update: v0.2.1b5 and v0.2.1b6" section.

---

Generated by the weekly article relevance audit workflow. Run ID: 23964749052

Generated by Article Relevance Audit — Weekly Health Check · ◷

[github-actions]

🤖 Auto-triage by issue-worker

Why this is suitable for Copilot:
This audit contains one persistent broken internal link (4th consecutive audit, confirmed still unfixed in codebase) with a one-line exact fix, plus 3 new cross-linking opportunities with precise suggested markdown text and file paths — all mechanical changes requiring no editorial judgment.

What Copilot should do:

1. Fix broken link in src/content/articles/agentic-ai-transforming-dev-teams.mdx — change:

   [GitHub Copilot CLI worktrees](/articles/copilot-fleet-mode-video-pipeline)
   ```
   to:
   ```
   [GitHub Copilot CLI worktrees](/articles/video-pipeline-with-fleet-mode)
   

   (The correct slug video-pipeline-with-fleet-mode exists. Located in the ## The Circle Replaces the Hierarchy section.)

2. Add cross-link in src/content/articles/copilot-cli-extensions-cookbook-examples.mdx — in the ### Example 16: The REPL Loop — Self-Healing Until Green section, insert after "all from a single prompt.":

   I took this pattern further in production — [I built a full CI monitor extension](/articles/ci-monitor-extension-agent-ci-feedback-loop) that wraps GitHub Actions into the loop, so the agent watches CI, reads failure logs, and iterates without me babysitting the terminal.

3. Add cross-link in src/content/articles/nvidia-openshell-sandbox-ai-agents.mdx — near the closing section on real-world applications, insert after "That's the part that changes everything: policy as code.":

   The real-world payoff of that security infrastructure is starting to show — [I built a Telegram bridge to Copilot CLI in ~420 lines](/articles/who-needs-openclaw-copilot-cli-extensions) that runs inside an OpenShell sandbox, using the L7 proxy for credential injection. Same agent, now remote-controlled from your phone.

4. Add cross-link in src/content/articles/copilot-cli-remote-access-your-agent-from-anywhere.mdx — near session continuity or remote access patterns, insert:

   I took this further by [building a Telegram bridge in a single extension file](/articles/who-needs-openclaw-copilot-cli-extensions) — no gateway, no daemon, just the extension SDK and a bot token. Full Copilot CLI, remote-controlled from your phone.

Confidence: High

Assigning to Copilot coding agent.

Generated by Issue Worker — Backlog Triage Agent · ◷