Who we’re following — intake list for Claude Code, 2026
Two reasons to keep up with a Claude Code source:
- Helps users — practical tips, configs, workflows, templates that make your next CC session better.
- Big news — releases, paradigm shifts, leaks, conference launches; anything notable enough to discuss broadly.
The list below is organised that way. Entries marked ★ are already cited in one of the six entries.
Same shape as the entries themselves: dated when possible, honest about what each source covers and doesn’t, no implicit ranking.
Helps users
Sources whose primary value is making your next session better — workflow recipes, templates, skill / plugin catalogs, debugging patterns, configs to copy.
Anthropic first-party — the user-facing surfaces
| Source | What | Where |
|---|---|---|
| ★ Claude Code docs | Reference for CLI / config / settings.json / MCP wiring | https://code.claude.com/docs/en/ |
| ★ Anthropic best-practices | Anthropic’s own “how to use Claude Code well” page — the canonical opinionated guide | https://code.claude.com/docs/en/best-practices |
| ★ Claude Cookbook | Worked examples for SDK + agent patterns; growing on every launch | https://platform.claude.com/cookbook/ |
Workflow demos + recipes
| Source | What |
|---|---|
| ★ Cole Medin (YouTube) | Long-form CC demos; the 24-hour-coding-agent video is a category-defining example — https://www.youtube.com/@ColeMedin |
| ★ Scott Spence | Practical write-ups on plugin marketplaces + skill organisation — https://scottspence.com/ |
| ★ Mark Chen (Medium) | Beginner-friendly walkthroughs; the skills-marketplace post is the cleanest intro to plugin install flow — https://medium.com/@markchen69 |
| ★ Wataru Takahashi | Spec-driven-development series; deep on the spec→plan→code workflow — https://watarutakahashi.substack.com/ |
| ★ DataCamp tutorials | Series on plan mode, TDD with Claude, spec-driven dev — https://www.datacamp.com/tutorial/spec-driven-development-with-claude-code |
| ★ firecrawl blog | “Best Claude Code Skills to Try in 2026” — useful for skill discovery — https://www.firecrawl.dev/blog/best-claude-code-skills |
| ★ BuildFastWithAI | The complete 2026 skills guide; tutorial-shaped — https://www.buildfastwithai.com/blogs/claude-skills-complete-guide-2026 |
| ★ Nimbalyst | Practical CC guides (skills + subagents) — https://nimbalyst.com/blog/ |
| Hamel Husain (Substack) | Practitioner-engineer commentary; intermittent but high-signal when he writes about CC |
| Eugene Yan — Importance of Being Earnest | ML/LLM-eng practitioner reflection; not CC-specific but the lens is right |
Templates + catalogs
| Source | What |
|---|---|
| ★ Forrest Chang’s CLAUDE.md | The 220K+ ⭐ fork-network anchor — the most-cited CLAUDE.md template in existence (works as starting point even if you write your own) |
| ★ anthropics/claude-plugins-official | The Anthropic-managed plugin marketplace — https://github.com/anthropics/claude-plugins-official |
| ★ agentskills.io | Cross-vendor open-standard skill index — https://www.agensi.io/learn/claude-code-plugin-marketplace-guide |
★ tonsofskills.com / jeremylongshore/claude-code-plugins-plus-skills |
Largest aggregator (425 plugins, 2,810 skills, 200 agents as of 2026-05) |
| claudemarketplaces.com | Cross-marketplace skill / plugin / MCP server directory; updates daily from GitHub |
★ netresearch/claude-code-marketplace |
Community-curated marketplace; agentskills.io reference implementation |
Practitioner conversations
| Source | What |
|---|---|
| ★ Latent Space (podcast + newsletter) | Highest-signal cross-vendor agentic-engineering channel; Swyx + Alessio gather the operators — https://latent.space/ |
| Latent Space Discord | Real-time practitioner channel where the workflow patterns get debated weeks before they hit blogs |
Big news
Sources whose primary value is catching what just happened — release notes, paradigm-shift posts, leak coverage, conference launches.
Official release channels
| Source | What | Cadence |
|---|---|---|
| ★ Anthropic Claude Code changelog | Per-version notes; canonical for hook events / flag changes / env vars | https://code.claude.com/docs/en/changelog — every 1–2 days during 2026 |
GitHub releases — anthropics/claude-code |
Raw release feed; arrives before the changelog page renders. RSS-able. | https://github.com/anthropics/claude-code/releases |
| Anthropic News / Product | Launch announcements (Code Review, Managed Agents, conference recaps) | https://www.anthropic.com/news |
| @AnthropicAI on X/Twitter | Same launches, faster surfacing during live events | twitter.com/AnthropicAI |
| Code w/ Claude conference | Anthropic’s annual developer event — first edition May 2026 (Code Review launch). Recorded talks land on YouTube after | https://www.anthropic.com/events + Anthropic YouTube |
| ★ Releasebot — Anthropic | Third-party automated changelog with diff-format excerpts; useful as a single feed across products | https://releasebot.io/updates/anthropic/claude-code |
| ClaudeLog third-party changelog | Editorial summaries on top of the raw changelog — useful when you want “what does this version mean for me” rather than the raw entries |
Practical note: Anthropic-side, the GitHub releases page arrives first, the changelog page catches up 24–48 hours later with prose. If you want “the moment a new version ships,” the GitHub releases RSS is the right feed; for the digested version, the changelog page.
Direction-of-travel voices (paradigm shifts, opinion-setting)
| Source | What |
|---|---|
| ★ Anthropic Engineering blog | The post-leak Effective context engineering for AI agents piece is the foundational 2026 framing of “prompt engineering is over, context engineering is here” — https://www.anthropic.com/engineering |
| ★ Boris Cherny (Head of Claude Code) | The most-cited single voice on CC in 2026; podcast circuit + @bcherny. The “coding is solved, what comes next” framing is his |
| ★ Cat Wu (founding engineer) | Engineering-side framing; tools & UX choices. Co-appears with Cherny on big interviews |
| ★ Simon Willison | Live-blogs Anthropic events; cited deep dives on CC behaviour — https://simonwillison.net/ |
| Andrej Karpathy | The “80% agent-driven coding” pivot post inspired the Forrest Chang CLAUDE.md pattern; not a regular CC commentator but his occasional pieces move the conversation |
Event coverage + investigative
| Source | What |
|---|---|
| ★ VILA-Lab — Dive into Claude Code | Academic analysis of the source leak; arxiv 2604.14228 + follow-up posts — https://github.com/VILA-Lab/Dive-into-Claude-Code |
| ★ Layer5 engineering blog | The single best deep-dive on the source leak (file counts, QueryEngine analysis) — https://layer5.io/blog/engineering/the-claude-code-source-leak-512000-lines-a-missing-npmignore-and-the-fastest-growing-repo-in-github-history/ |
| ★ Cognition Labs blog | Cross-ecosystem big-news items (the claude-memory-compiler post was the first serious Karpathy-wiki implementation for CC) — https://www.cognitionus.com/blog |
Podcast circuit (where Anthropic’s framing gets articulated)
| Show | Hosts | Why listen |
|---|---|---|
| ★ The Pragmatic Engineer | Gergely Orosz | Cherny’s March 2026 appearance is the most cited single CC interview — engineering-management lens |
| ★ Every / AI & I | Dan Shipper | Cherny + Wu joint appearance (October 2025) was the foundational “how we use CC inside Anthropic” interview |
| ★ Lenny’s Newsletter Podcast | Lenny Rachitsky | Cherny “what comes after coding is solved” — product-management lens |
| No Priors | Sarah Guo + Elad Gil | Anthropic-adjacent strategy; occasional CC-relevant interviews |
Peripheral — adjacent ecosystems, when relevant
These don’t fit the “helps users” or “big news” buckets directly — they’re parallel-evolution agent tools whose movements occasionally predict where CC goes or carry techniques worth borrowing. Watch when actively relevant; don’t follow on default cadence.
| Source | What | Why peripheral-watch |
|---|---|---|
| ★ Paul Gauthier (Aider) | https://aider.chat/ | TDD-with-AI patterns matured here first |
| Cursor blog | https://cursor.com/blog | Their rules system is the spiritual sibling of CLAUDE.md |
| Continue.dev blog | https://continue.dev/blog | Open-source IDE-agent; useful for cross-surface skill comparison |
| OpenAI Codex CLI / Gemini CLI / GitHub Copilot Agent mode | Respective vendor docs | Cross-vendor skill compatibility starts here; agentskills.io spans all of them |
What we’re not yet watching (gaps to fill)
Honest about the holes:
- CN-language CC writing. Substantial Chinese-language CC content on Bilibili / Zhihu / WeChat that we’re not pulling from yet.
- JP-language CC writing. Less surface area but high-signal — the Japanese dev community has been early on agent patterns.
- Discord channel notes. The Latent Space / Anthropic Discords carry real conversation that doesn’t make it to blog form for weeks. No discipline yet for distilling those.
- Specific operator threads on X — Karpathy, Yi Tay, Hamel Husain. Cite when load-bearing; don’t follow systematically yet.
- Enterprise / private channels. Anthropic Slack channels for enterprise customers have earlier signal than public surfaces. Not applicable at TinySmile scale today.
How we update this file
Same rules as the entries: add new sources by appending to the relevant section; mark with ★ once we’ve cited them in an entry. When a source goes quiet for >6 months, move it to a “former” section rather than deleting — readers should see the ecosystem’s churn, not just its current state.
When a source publishes something we want to cite, the workflow:
- Add the link inline at the point of claim in the relevant entry
- Make sure the source appears in this file with the ★ marker
- If new to this file, add it under the right section with the “what to expect” line
Discipline: don’t add a source here unless we’d actually use it. The list earns its place through use, not aspirational following.