Building AI doc toolchains with Claude skills

I’ve been building out some powerful AI-assisted docs toolchain infrastructure recently. Some random observations:

• Claude’s plugins are amazing! Especially the official /skill-creator and /code-simplifier plugins. I need to check out the official brainstorming skill next.
• You know how the term “task switching” is in our workplace lexicon? I want to add “pace switching” too. It’s the feeling of working with AI on something blazingly fast, and then abruptly slowing down to carefully check and correct its work. It’s like stumbling off a moving walkway sometimes.
• LLM coding agents are SUCH a loss leader right now – like a $200/month subscription could use up to $2,000 in compute. Designing my AI toolchains to minimize token use seems pretty wise, and I’m already finding ways I can steer LLMs in that direction. Like, “hey, agent, go use the llms.txt to search the docs, and then fetch the MD raw files. Don’t just use our sensible-docs MCP server for docs search!” Then again, I could be wrong, and LLM token use will get so efficient and so cheap that it won’t be a concern.
• The AI docs-authoring toolchain I’ve built out for my client Sensible is pretty extensive at this point – my .claude dir is full of docs-writing skills, preferences, style guides, and templates. And the crazy thing is, they really do seem to work well! I’m growing and changing the infrastructure fast – this snapshot will probably be out of date in half a week:

.claude/
├── preferences/
│   └── editorial-preferences.md
└── skills/
    ├── create-concept-doc/
    ├── create-docs-example/
    ├── create-new-doc/
    ├── revise-doc-style/
    ├── sensible-changelog/
    ├── sensible-integration-guide-gen.../
    ├── sensible-style-guide-generator.../
    ├── update-docs-from-pr/
    ├── update-existing-doc/
    └── upload-pr-extractor/

style-guide/
├── changelog-style-guide.md
├── concept-topic-template.md
├── config-library-supported-docu.../
└── integration-guide-template.md

Next up: check out a skill-validator whose author claims the output of Claude’s skill-creator often actually fails her validation tool!