When a Tool Result Is Too Big and Melts Your Context Window: Designing Cursor-Based Pagination
When a list tool returns hundreds of rows at once, an agent's context can collapse in a single call. Here is a cursor-based pagination design that keeps tool output small and protects your token budget, with working code.
Did That Post Actually Go Through? Safely Retrying an Interrupted MCP Write Without Double-Executing
When an MCP write tool call is interrupted by a dropped connection, you can't tell whether the server ran it. Here's why naive retries cause double-execution, and a working wrapper that uses idempotency keys and a reconcile read to retry safely — with examples from an unattended pipeline.
When an OAuth Token Expires, Your Unattended Run Has Nowhere to Go — A Token-Lifecycle Design That Keeps Remote MCP Alive
Remote MCP connectors are authorized via OAuth, but access tokens are short-lived. Interactive sessions can re-authorize in a browser; an unattended scheduled run has nobody to click the dialog. Here is a token-lifecycle design that owns expiry and refreshes ahead of time.
The Day a Non-Responding MCP Call Swallowed an Entire Unattended Run — Owning the Stop With Your Own Deadline
When a remote MCP tool call stops responding, an unattended scheduled run just keeps waiting. Instead of leaving the cutoff entirely to the platform, here is how I designed my own deadline and a per-connector circuit breaker to own the stop — with working code.
Reach a Remote MCP Server in a Single API Request: Implementing the Messages API MCP Connector
How to call a remote MCP server's tools using only the Messages API's mcp_servers and mcp_toolset—no local MCP client. Covers allowlist/denylist design, response handling, and the pitfalls to avoid before unattended production use.
Connecting Managed Agents to Services You Don't Want to Expose: MCP Tunnel Design
How to connect Claude Managed Agents to an internal MCP server that is never exposed to the public internet. We cover the MCP tunnel, self-hosted sandboxes, authorization boundaries, and graceful degradation when things break.
When Claude Code Still Writes Stale Code With Context7 Installed — Verifying the Injection Actually Worked
You installed Context7 MCP, yet Claude returns code against a deprecated API. Most of the time the documentation injection silently no-ops. Here are field notes on a hook that detects misses, pinning doc versions, and a harness that checks generated code every time.
Keep MCP Connector Authorization in One Place — A Solo Dev Design That Survives Growing Integrations
When Claude chat, Claude Code, and Cowork each configure the same MCP connector separately, their authorization drifts apart and breaks silently. Here is how to borrow the managed-connector idea of 'provision once, reuse everywhere' as an indie developer.
Give an Unattended Agent Only the MCP Tools It Needs — Enforcing a Deny-by-Default Policy
An unattended Claude Code agent can't lean on a permission prompt, so whatever a tool can reach becomes the blast radius. Here's how to lock MCP servers and tools down to deny-by-default and hand back only what the job needs, with managed-settings.json examples.
Finding and Installing Claude Code Plugins from a Marketplace — Read What's Inside Before You Trust It
A practical walkthrough for discovering and installing Claude Code plugins from a marketplace. Beyond the /plugin commands, it covers how to read the bundled contents before you install, and how to choose between User and Project scope.
Beyond Tools in MCP: Designing with Resources, Prompts, and Sampling
Cramming everything into MCP tools hits a wall fast. Here is how resources, prompts, and sampling untangle a server, told through a real wallpaper-app asset manager I cut from 14 tools down to 5.
Wiring Chrome DevTools for agents 1.0 into Claude Code: a setup memo
Google shipped Chrome DevTools for agents 1.0 as a stable release. Here is how I wired it into Claude Code, why I picked the MCP path over the CLI, and how it compares to the version bundled in Antigravity 2.0.