Fix: .mcp.json servers never start because the workspace is marked untrusted
Project MCP servers stuck at Pending approval are not misconfigured. Run claude interactively in the folder, accept the workspace trust dialog, then approve the servers.
Tag: ai-agents
166 posts · Page 4 of 17
Visual Studio 18.9 Lets You Set Thinking Effort Per Model
Visual Studio 18.9 Insiders 2 adds a per-model thinking effort control with Low through Max levels, exposing the same reasoning-effort dial the underlying model APIs take.
Fix: a Windows path in your agent config silently turns into a tab, a newline, or a Chinese character
Folder names starting with t, b, n, r or u make a Windows path parse cleanly as JSON and still resolve wrong. Use forward slashes or double every backslash.
MCP C# SDK 2.0 Ships: Stateless by Default and MCP9005 on Your Old Code
ModelContextProtocol 2.0.0 landed on 2026-07-28 with the stateless HTTP transport on by default, Multi Round-Trip Requests replacing server-initiated elicitation, and an analyzer warning on ElicitAsync and SampleAsync.
GitHub's MCP Server Went Stateless and Deleted Its Redis Session Store
On 2026-07-23 GitHub shipped the MCP 2026-07-28 revision ahead of the spec date. The notable part is the subtraction: no initialize handshake, no Mcp-Session-Id, no Redis.
How to Run a Background Coding Agent That Auto-Commits and Opens a Draft PR When It Finishes
Two ways to make a coding agent land its own work: Claude Code 2.1.198+ background sessions commit, push, and open a draft PR from their worktree with no config, or wrap any agent CLI in a git worktree plus gh pr create --draft. Covers the worktree guardrails, the subagent-vs-session distinction, the GITHUB_TOKEN trigger trap, and why there is still no off switch.
Cursor Router Makes Auto a Per-Request Model Decision
Cursor Router shipped on July 22, 2026. Auto now classifies every request and routes it to a different model, and the Cost, Balance, and Intelligence modes change both the quality you get and how you are billed.
How to Stream Nested Subagent Output From a Headless Claude Code Run
By default a headless Claude Code run emits only tool_use and tool_result blocks from its subagents, so the delegated reasoning is invisible. Pass --forward-subagent-text (Claude Code 2.1.211+) to get subagent text and thinking in the stream-json output, and demultiplex it by parent_tool_use_id.
Claude Code 2.1.219 Reopens Nested Subagents, Three Layers Deep
Version 2.1.219 raises the default subagent spawn depth from 1 to 3, adds a workflowSizeGuideline settings key, and ships a fail-closed sandbox network allowlist.
How to Lock Down a Coding Agent's Network Egress With a Strict Host Allowlist
Claude Code, Cursor, and the GitHub Copilot coding agent all ship host allowlists for outbound traffic, and all three default to something looser than you want. The exact settings keys, a policy you can copy, and the four surfaces the allowlist does not cover.