Claude Productivity Tools: The Features Most People Never Find
Most people use maybe 30% of Claude: they open a chat, type a question, copy the answer. The productivity gains everyone talks about live in the other 70% — the features that turn one-off conversations into systems that remember your context, produce real artifacts, and connect to where your work actually lives. Here's the tour, roughly in the order the upgrades pay off.
1. Projects: stop re-explaining yourself
A Project is a workspace with persistent context: upload your docs, define custom instructions, and every conversation inside it starts already knowing your company, your style guide, your codebase conventions. One Project per ongoing concern — your product, your course, your client — is the single biggest quality-of-life upgrade in the Claude apps. The pattern: the first ten minutes you spend loading a Project with context repay themselves in every conversation after.
2. Artifacts: outputs you can actually use
Ask for a document, a diagram, an interactive calculator, or a working mini-app, and Claude builds it as an artifact — a live, editable thing beside the chat rather than a wall of text to copy. Underrated uses: interactive mockups for stakeholder feedback, one-off internal tools (rota generators, calculators), and formatted documents that survive contact with a real audience.
3. File analysis: the drag-and-drop analyst
PDFs, spreadsheets, images, exports — drop them in and interrogate them. Claude's analysis tooling can run code over data files, which upgrades "summarize this CSV" to "compute the actual cohort retention in this CSV." For the technique layer, our data analysis prompts are built for exactly this.
4. Connectors: Claude where your work lives
Through MCP connectors, the Claude apps reach your other tools — Drive and calendars, Slack, Notion, project trackers — so "summarize this week's activity in the project channel and draft Monday's update" is one request, not an export-paste-prompt relay. Start with one connector you'd use daily; our MCP guide covers choosing safely.
5. Skills: teach it your way of working
Skills — folders of instructions Claude loads on demand — began in Claude Code but now power specialized behavior across Claude surfaces: document formats, review checklists, domain workflows. The practical effect: instead of writing a 400-word prompt every time, you invoke a capability that already knows the standard. Write your own (lint it with our free SKILL.md validator) or install a library.
6. Claude Code — not just for engineers
The secret spreading through non-engineering teams: Claude Code is a general file agent. Marketers run content pipelines with it, analysts batch-process spreadsheets, founders automate reporting — anything that reads and writes files. If your work produces artifacts in folders, the terminal is worth the fear.
7. The habit layer: what power users do differently
- They maintain context once — Projects and CLAUDE.md files instead of re-pasting background every session.
- They save winning prompts — a personal library (or ours: 50 free prompts) beats re-improvising.
- They chain instead of cram — research → outline → draft → edit as separate steps with review between.
- They verify at the boundary — anything leaving the building (numbers, claims, citations) gets checked. Speed without this habit is just faster mistakes.
The productized version: the ClaudeThings kits install 89 agents, 103 skills, and 181 slash commands — engineering and marketing workflows that turn this whole page into a one-command setup. See the kits →