Mcp Servers | Techlist.io

githubApr 30, 20262 min readCurated summary

GitHub Copilot CLI for Beginners: Interactive v. non-interactive mode

GitHub Copilot CLI offers two ways to work from the terminal: interactive mode for ongoing, collaborative sessions and non-interactive mode for quick, one-off prompts. The article recommends choosing between them based on whether you need iterative exploration or a fast answer within an existing shell workflow. Previous sessions can also be resumed in either mode. ## Interactive Mode for Collaborative Work - Start it by running `copilot`. - It is the default mode and provides a chat-like, back-and-forth experience. - Users can: - Ask questions about a project. - Review Copilot’s responses. - Ask follow-up questions. - Request actions, such as running a local server. - Copilot may request permission to trust the current folder because it needs to read or modify project files. - This mode is best for exploratory coding, troubleshooting, and tasks that require multiple steps. ## Non-Interactive Mode for Quick Prompts - Start it with `copilot -p "your prompt"`. - It produces a single response without opening a full conversational session. - Typical uses include: - Summarizing a repository. - Generating code snippets. - Running Copilot inside automated workflows. - It is designed to keep users in their normal terminal flow and is most useful when the required task is clear and narrowly defined. ## Resuming Previous Sessions - In interactive mode, enter `/resume` to select a previous session. - From the regular shell, use `copilot --resume` to open the session picker directly. - Resuming preserves the context of earlier conversations, making it easier to continue unfinished work. The practical choice is simple: use interactive mode for deeper, iterative collaboration and non-interactive mode for focused, one-shot requests.

Read original

Mcp Servers Github Copilot Cli Command Line Interface Interactive Mode+2

figmaSep 9, 20252 min readCurated summary

Issue No.12: New Roles, New Rules | Figma Blog

Figma’s “New roles, new rules” highlights how AI and faster iteration are blurring traditional boundaries between product roles. Product managers, designers, and developers are increasingly working across disciplines, using prototypes and shared principles to collaborate directly. The issue argues that effective teams are replacing rigid handoffs with experimentation, co-creation, and better systems for guiding AI. ## Shifting Roles - Research found that: - 64% of product builders identify with two or more roles. - 56% of non-designers perform design-related work. - Faster development cycles and AI tools are enabling people to contribute further outside their formal specialties. - As responsibilities expand, teams are also reconsidering how they manage time, ownership, and collaboration. ## Prototyping to Create Shared Understanding - Figma product designer Natasha Tenggoro struggled to explain how video playback should work in Figma Buzz. - Instead of relying on verbal descriptions, she used Figma Make to build prototypes herself. - The prototypes helped the team reach three “aha” moments and provided a clearer basis for discussion. - The example demonstrates how building interactive artifacts can replace lengthy explanations and accelerate alignment. ## Music-Inspired Design in Figma Draw - Figma Draw’s new scatter brushes were shaped by musical concepts such as tempo, texture, and volume. - Designers translated the character of genres including Honky-tonk, Screamo, Doo-wop, and Vaporwave into brush behavior. - The release adds 10 scatter brushes, giving designers more control over attributes such as gap, wiggle, and jitter. ## Duolingo’s Collaborative Method - Duolingo’s Math team is changing the traditional design-to-engineering handoff for its math games. - Designers and engineers work through co-creation, scrappy prototypes, and continuous experimentation. - Shared principles—including “show, don’t tell” and distinguishing between a v1 and an MVP—help the team move quickly while refining ideas. - Collaboration is treated as an ongoing product practice rather than a stage that ends when design is handed off. ## Broader Changes in Creative Work - Agencies and freelancers are also abandoning rigid client boundaries and involving clients throughout the creative process. - Design systems can improve AI-generated code by giving agents structured, relevant, and brand-consistent input. - MCP servers are presented as an important connection between design systems and AI-powered workflows, helping agents produce more useful output. Overall, the issue recommends embracing broader roles and replacing formal handoffs with shared prototypes, collaborative iteration, and well-structured design systems—especially as AI becomes more involved in product development.

Read original

Mcp Servers Figma AI Design Systems+2