Skip to content

The Deska blog

Blog

Product notes, release highlights and how we build Deska: the one infinite canvas workspace for your editor, terminal, browser, notes and AI agents.

  1. 10 min readThe Agent Hardcoded Something It Should Not HaveLearn why the agent hardcoded a value during your build and how to use multi-agent observability to catch these silent failures before deployment.
  2. 10 min readHanding a Task From One Model to AnotherLearn how to optimize workflows by handing a task from one model to another using tiered architectures and specialized agent environments.
  3. 11 min readThe Agent Invented an API: Catching Hallucinated Code EarlyLearn why AI agents create non-existent libraries and how to catch hallucinated code using local execution, terminal feedback, and side-by-side verification.
  4. 10 min readGraphical Frontends for Coding AgentsExplore the top graphical frontends for coding agents in 2026. Compare IDE extensions, standalone canvas apps, and local-first developer tools like Deska.
  5. 11 min readMigrating Off GraphQL With an AgentLearn how to execute a successful GraphQL to REST migration using AI agents to map schemas, generate endpoints, and update client logic efficiently.
  6. 12 min readBreaking Up a God Object, Method by MethodLearn how to start breaking up a God Object through controlled, method-level refactoring to improve software maintainability and reduce technical debt.
  7. 12 min readRefactoring a Spaghetti Workflow File With an AgentLearn how to refactor a spaghetti workflow file with an agent to improve CI performance, maintainability, and security in your GitHub Actions pipelines.
  8. 10 min readHow Agents Behave With GitExplore how agents behave with Git, focusing on commit quality, branch management, and how Deska integrates tools like Claude Code and Codex CLI.
  9. 10 min readGenerating Fixtures and Sample Data with AI AgentsLearn how to use AI coding agents for generating fixtures and sample data to streamline development, improve test coverage, and automate repetitive tasks.
  10. 10 min readBuilding a GDPR Data Inventory With Agent LegworkLearn how to build a GDPR data inventory using AI agents. This guide covers technical recipes for mapping personal data across local and cloud environments.
  11. 10 min readAdding Full-Text Search Without a Search ServerLearn methods for adding full-text search to local-first apps without a search server, utilizing client-side libraries and local indexing strategies.
  12. 10 min readMaking a Codebase Agent-Friendly: Structure That Helps AI Help YouLearn how to optimize your project for an agent-friendly codebase using documentation, modularity, and tools like Deska to improve AI coding efficiency.
  13. 11 min readAgent Framework or Your Own Loop?Deciding between an agent framework or your own loop involves weighing abstractions against control. Learn which architecture fits your AI coding project.
  14. 12 min readFramework Migrations With AI Agents: A Survival GuideLearn how to navigate complex framework migrations with AI agents by utilizing multi-agent strategies and local-first developer tools like Deska.
  15. 10 min readCode Archaeology With Agents: Understanding Inherited SystemsLearn how to perform code archaeology with agents to map legacy systems, trace logic, and document inherited codebases efficiently using modern AI tools.
  16. 10 min readUsing Agents Only for Boilerplate: A Valid StrategyExplore why using agents only for boilerplate is a valid strategy for developers who want to maintain control while accelerating repetitive setup tasks.
  17. 10 min readA Fleet of Agents for a Team of OneLearn how a fleet of agents for a team of one can multiply developer productivity through parallel execution and structured workspace management.
  18. 10 min readObservability for an Agent FleetMaster observability for an agent fleet. Learn how to monitor, debug, and orchestrate multiple AI agents in a local-first development environment.
  19. 10 min readCI Fails Only on Tuesdays: An Agent InvestigationLearn how to conduct a CI investigation for flaky tests using specialized tools and AI agents to solve non-deterministic build failures.
  20. 10 min readHunting Flaky Tests With an AI AgentLearn how to use an AI agent to identify and fix flaky tests by automating trace analysis and environmental debugging in a unified workspace.