an open-source alternative to the dots, bots and muses of the world, run an office of claude code/codex like agents on your laptop, sandboxes or anywhere, uses your existing subscriptions
HTML agent-harnessagent-orchestrationagentsai-agentsautonomous-agentsclaude-codecodexcoding-agentsdesktop-appdeveloper-toolselectrongemini-cliharnessharness-engineeringlocal-firstmemorymulti-agentopencodeorchestrationtypescript
blog(mission): 3 new posts from the first mission run
chaitanyagirif8717430a137
diff --git a/blog/media-src/spec.js b/blog/media-src/spec.jsindex f27f7f0c..f88b5c7c 100644--- a/blog/media-src/spec.js+++ b/blog/media-src/spec.js@@ -25,6 +25,9 @@ const SPEC = {"claude-squad-vs-munder-difflin": { a: "versus", blue: "claude squad", blue2: "munder difflin", orange: "which should you use?" },"cline-vs-munder-difflin": { a: "versus", blue: "in-editor agent", blue2: "local hive", orange: "cline vs munder difflin" },"command-center-guide": { a: "spotlight", blue: "kanban · fleet · budgets", orange: "one screen for the whole floor" },+ "how-to-add-an-mcp-server-to-claude-code": { a: "wire", blue: "claude mcp add", orange: "one command, three scopes" },+ "codex-cli-vs-claude-code": { a: "versus", blue: "codex cli", blue2: "claude code", orange: "same job, two dials" },+ "how-much-does-claude-code-cost": { a: "ledger", blue: "pro · max · api", orange: "what the bill really says" },"compressing-agent-memory": { a: "memory", blue: "compress, keep the original", orange: "smaller memory, no loss" },"conductor-claude-code-alternative": { a: "versus", blue: "conductor", blue2: "munder difflin", orange: "a conductor alternative" },"context-engineering-for-ai-agents": { a: "ledger", blue: "the context window is a budget", orange: "spend it deliberately" },diff --git a/blog/src/_data/media.json b/blog/src/_data/media.jsonindex f593b841..85ead832 100644--- a/blog/src/_data/media.json+++ b/blog/src/_data/media.json@@ -3407,5 +3407,80 @@"status": "placeholder"}]+ },+ "how-to-add-an-mcp-server-to-claude-code": {+ "title": "How to Add an MCP Server to Claude Code",+ "category": "guides",+ "hero": {+ "file": "assets/media/how-to-add-an-mcp-server-to-claude-code/hero.png",+ "alt": "Illustration: How to Add an MCP Server to Claude Code",+ "prompt": "Hero image for a blog post titled \"How to Add an MCP Server to Claude Code\". The post is about: The claude mcp add command, its scope flag, editing .mcp.json by hand, and how Munder Difflin's Capabilities screen grants an MCP server to one agent at a time.",+ "status": "ready"+ },+ "inline": {+ "note-1": {+ "file": "assets/media/how-to-add-an-mcp-server-to-claude-code/note-1.png",+ "alt": "Hand-drawn sketch from “How to Add an MCP Server to Claude Code”",+ "prompt": "",+ "status": "ready"+ },+ "note-2": {+ "file": "assets/media/how-to-add-an-mcp-server-to-claude-code/note-2.png",+ "alt": "Hand-drawn sketch from “How to Add an MCP Server to Claude Code”",+ "prompt": "",+ "status": "ready"+ }+ },+ "youtube": []+ },+ "codex-cli-vs-claude-code": {+ "title": "Codex CLI vs Claude Code",+ "category": "comparisons",+ "hero": {+ "file": "assets/media/codex-cli-vs-claude-code/hero.png",+ "alt": "Illustration: Codex CLI vs Claude Code",+ "prompt": "Hero image for a blog post titled \"Codex CLI vs Claude Code\". The post is about: Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.",+ "status": "ready"+ },+ "inline": {+ "note-1": {+ "file": "assets/media/codex-cli-vs-claude-code/note-1.png",+ "alt": "Hand-drawn sketch from “Codex CLI vs Claude Code”",+ "prompt": "",+ "status": "ready"+ },+ "note-2": {+ "file": "assets/media/codex-cli-vs-claude-code/note-2.png",+ "alt": "Hand-drawn sketch from “Codex CLI vs Claude Code”",+ "prompt": "",+ "status": "ready"+ }+ },+ "youtube": []+ },+ "how-much-does-claude-code-cost": {+ "title": "How Much Does Claude Code Cost?",+ "category": "concepts",+ "hero": {+ "file": "assets/media/how-much-does-claude-code-cost/hero.png",+ "alt": "Illustration: How Much Does Claude Code Cost?",+ "prompt": "Hero image for a blog post titled \"How Much Does Claude Code Cost?\". The post is about: What Claude Code actually costs: free plan access, Claude Pro and Max pricing, per token API rates, and what running several agents does to the bill.",+ "status": "ready"+ },+ "inline": {+ "note-1": {+ "file": "assets/media/how-much-does-claude-code-cost/note-1.png",+ "alt": "Hand-drawn sketch from “How Much Does Claude Code Cost?”",+ "prompt": "",+ "status": "ready"+ },+ "note-2": {+ "file": "assets/media/how-much-does-claude-code-cost/note-2.png",+ "alt": "Hand-drawn sketch from “How Much Does Claude Code Cost?”",+ "prompt": "",+ "status": "ready"+ }+ },+ "youtube": []}}diff --git a/blog/src/assets/media/codex-cli-vs-claude-code/hero.png b/blog/src/assets/media/codex-cli-vs-claude-code/hero.pngnew file mode 100644index 00000000..cdeec885Binary files /dev/null and b/blog/src/assets/media/codex-cli-vs-claude-code/hero.png differdiff --git a/blog/src/assets/media/codex-cli-vs-claude-code/note-1.png b/blog/src/assets/media/codex-cli-vs-claude-code/note-1.pngnew file mode 100644index 00000000..14f02337Binary files /dev/null and b/blog/src/assets/media/codex-cli-vs-claude-code/note-1.png differdiff --git a/blog/src/assets/media/codex-cli-vs-claude-code/note-2.png b/blog/src/assets/media/codex-cli-vs-claude-code/note-2.pngnew file mode 100644index 00000000..1c1abb31Binary files /dev/null and b/blog/src/assets/media/codex-cli-vs-claude-code/note-2.png differdiff --git a/blog/src/assets/media/how-much-does-claude-code-cost/hero.png b/blog/src/assets/media/how-much-does-claude-code-cost/hero.pngnew file mode 100644index 00000000..e34c0eabBinary files /dev/null and b/blog/src/assets/media/how-much-does-claude-code-cost/hero.png differdiff --git a/blog/src/assets/media/how-much-does-claude-code-cost/note-1.png b/blog/src/assets/media/how-much-does-claude-code-cost/note-1.pngnew file mode 100644index 00000000..b080349bBinary files /dev/null and b/blog/src/assets/media/how-much-does-claude-code-cost/note-1.png differdiff --git a/blog/src/assets/media/how-much-does-claude-code-cost/note-2.png b/blog/src/assets/media/how-much-does-claude-code-cost/note-2.pngnew file mode 100644index 00000000..e0182334Binary files /dev/null and b/blog/src/assets/media/how-much-does-claude-code-cost/note-2.png differdiff --git a/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/hero.png b/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/hero.pngnew file mode 100644index 00000000..d6e4e9fdBinary files /dev/null and b/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/hero.png differdiff --git a/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/note-1.png b/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/note-1.pngnew file mode 100644index 00000000..fd7a4a24Binary files /dev/null and b/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/note-1.png differdiff --git a/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/note-2.png b/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/note-2.pngnew file mode 100644index 00000000..7cfe6e64Binary files /dev/null and b/blog/src/assets/media/how-to-add-an-mcp-server-to-claude-code/note-2.png differdiff --git a/blog/src/posts/best-ai-coding-agents.md b/blog/src/posts/best-ai-coding-agents.mdindex 69c099c2..e318563a 100644--- a/blog/src/posts/best-ai-coding-agents.md+++ b/blog/src/posts/best-ai-coding-agents.md@@ -112,7 +112,7 @@ enough. If that's your need specifically with Claude Code, theFor a structured rubric across all of these, see [how to choose a multi-agent tool](/blog/how-to-choose-a-multi-agent-tool/);for two head-to-heads, [Cline vs](/blog/cline-vs-munder-difflin/) and-[Claude Squad vs Munder Difflin](/blog/claude-squad-vs-munder-difflin/).+[Claude Squad vs Munder Difflin](/blog/claude-squad-vs-munder-difflin/). Choosing between two terminal agents? [Codex CLI vs Claude Code](/blog/codex-cli-vs-claude-code/) compares them head to head.## The bottom linediff --git a/blog/src/posts/codex-cli-vs-claude-code.md b/blog/src/posts/codex-cli-vs-claude-code.mdnew file mode 100644index 00000000..d7ebcfe5--- /dev/null+++ b/blog/src/posts/codex-cli-vs-claude-code.md@@ -0,0 +1,67 @@+---+title: "Codex CLI vs Claude Code"+description: "Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins."+date: 2026-09-10+category: comparisons+categoryLabel: Comparisons+type: Technical+primaryKeyword: "codex cli vs claude code"+secondaryKeywords: ["claude code vs codex cli", "codex cli vs claude code comparison", "codex cli approval mode", "codex cli permission mode", "does codex cli use agents md", "codex cli vs claude code reddit"]+tags: ["Comparisons", "Claude Code", "Codex", "Multi-Agent", "Engines"]+faq:+ - q: "Is Codex CLI free to use?"+ a: "There is a real free tier, but it is capped. OpenAI's own Codex pricing page lists a Free plan at $0 a month for quick coding tasks, a Go plan at $8, and a Plus plan at $20 that adds cloud-based runs and the GPT-5.6 model family, checked 10 Sep 2026. Claude Code has no equivalent free tier: it authenticates against a paid Claude subscription or an Anthropic API key."+ - q: "Does Codex CLI read AGENTS.md the way Claude Code reads CLAUDE.md?"+ a: "Yes, but they are separate files, and each CLI's docs describe only its own. Codex CLI's docs list /init as the command that creates an AGENTS.md with instructions for Codex, checked 10 Sep 2026. Claude Code instead reads a project's CLAUDE.md at the start of every session, per Anthropic's own docs checked the same day, and neither doc promises to read the other's file."+ - q: "What is the real difference between Codex CLI's sandbox and Claude Code's permission mode?"+ a: "Codex CLI splits the decision in two: --sandbox (read-only, workspace-write, or danger-full-access) decides what it can touch, and --ask-for-approval (on-request or never) decides when it has to stop and ask, confirmed against codex --help on 10 Sep 2026. Claude Code folds both into one --permission-mode flag with six named modes, confirmed against claude --help the same day. Neither approach is wrong, Codex's is more composable and Claude Code's is a single dial to reason about."+ - q: "Can I run Codex CLI and Claude Code together?"+ a: "Yes. Munder Difflin wraps both as real terminal processes and runs Codex and Claude Code agents side by side on the same office floor, so each can take its own task at the same time. Munder Difflin 0.5.0 (2026-09-08) added GPT-6 Astra to Codex's model picker, and that slug needs Codex 0.153.1 or newer to be accepted when the agent spawns."+ - q: "Codex CLI vs Claude Code, which one is actually better?"+ a: "Neither wins outright. Codex CLI is the better pick when you want an open source binary you can audit, a free way in, and separate dials for sandboxing and approval. Claude Code is the better pick when you already pay for Claude and want a single permission mode, CLAUDE.md, and hooks that run around every tool call."+---++Pick Codex CLI if you want an open source agent with a free plan and separate dials for sandboxing and approval. Pick Claude Code if you already pay for Claude and want one permission mode, `CLAUDE.md` and hooks around every tool call. Both run real commands in your repo; the difference is how each hands you control.++Both are real terminal agents from frontier labs, and the differences that matter show up in how each one is built and how it hands you control, not in vague claims about which model is smarter. It's the same lifecycle question we asked when comparing [GitHub Copilot CLI against Claude Code as hive workers](/blog/github-copilot-cli-vs-claude-code-for-hive-work/): which engine stays resident, and which one behaves the same way the second time you hand it control. If you're weighing whole categories of [multi-agent tooling built around Claude Code](/blog/best-claude-code-multi-agent-tools/) rather than just these two engines, start there instead. This post stays narrow: Codex CLI against Claude Code, checked on 10 Sep 2026.++## What's the real difference between Codex CLI and Claude Code?++Codex CLI is an open-source, Rust-based binary from OpenAI that wraps each shell command in an OS-level sandbox and asks for approval on a separate policy. Claude Code is Anthropic's CLI, and it folds sandboxing and approval into a single permission-mode setting instead of two.++We checked this ourselves on 10 Sep 2026, with both CLIs already installed (`command -v claude` and `command -v codex` both resolved). `codex --help` lists `-s, --sandbox <SANDBOX_MODE>` with three possible values (`read-only`, `workspace-write`, `danger-full-access`), and a separate `-a, --ask-for-approval <APPROVAL_POLICY>` (`on-request` or `never`). `claude --help` lists one `--permission-mode <mode>` flag instead, with six named modes: `acceptEdits`, `auto`, `bypassPermissions`, `manual`, `dontAsk`, `plan`. Same underlying job, two different shapes: Codex separates "what it's allowed to touch" from "when it must ask," Claude Code bundles both into one choice.++OpenAI's own docs describe `workspace-write` as the mode where Codex "can read files, make edits, and run commands in the workspace," and `on-request`, the policy the CLI runs by default, as the one where Codex "requires approval to edit outside the workspace or to access network" and otherwise carries on ([developers.openai.com/codex/agent-approvals-security](https://developers.openai.com/codex/agent-approvals-security), checked 10 Sep 2026). Both projects also named their full escape hatch the same way: Codex's is `--dangerously-bypass-approvals-and-sandbox`, Claude Code's is `--dangerously-skip-permissions`. Somebody in both API design rooms clearly wanted you to feel that flag before you typed it, which is about as close as a CLI flag gets to a performance review.++## Does Codex CLI read `AGENTS.md` the way Claude Code reads `CLAUDE.md`?++Yes, but the two files are separate, and each CLI's docs describe only its own. Codex CLI's docs list `/init` as the command to "create an `AGENTS.md` file with instructions for Codex" ([developers.openai.com/codex/cli](https://developers.openai.com/codex/cli/), checked 10 Sep 2026). Claude Code instead reads a project's `CLAUDE.md` at the start of every session, described by Anthropic as "a markdown file you add to your project root that Claude Code reads at the start of every session" ([code.claude.com/docs/en/overview](https://code.claude.com/docs/en/overview), checked 10 Sep 2026).++The two overlap heavily in practice: coding standards, architecture notes, a review checklist. What differs is delivery. Neither CLI's docs promise to read the other's file, so a repo that runs both engines usually keeps both `AGENTS.md` and `CLAUDE.md`, or keeps one and symlinks the other to it.++{% img "note-1" %}++## Is Codex CLI free to use?++There's a real free tier, but it's capped. OpenAI's Codex pricing page lists a Free plan at $0 a month for exploring Codex "on quick coding tasks," a Go plan at $8 a month, and a Plus plan at $20 a month that adds cloud-based runs and the GPT-5.6 model family ([developers.openai.com/codex/pricing](https://developers.openai.com/codex/pricing), checked 10 Sep 2026). Claude Code has no equivalent free tier of its own: it authenticates against a paid Claude subscription or an Anthropic API key from the first session.++## Where does Codex CLI actually win?++Two places: openness and cost of entry. The CLI itself is Apache-2.0 licensed and public on GitHub, so you can read exactly what the sandbox and approval code does instead of trusting a closed binary ([github.com/openai/codex](https://github.com/openai/codex), checked 10 Sep 2026), and the free plan above gets a curious developer to a real sandboxed agent at no cost. If you want a run that never pauses to ask, `--ask-for-approval never` is also a cleaner single flag than composing Claude Code's mode and its skip flag together. None of that makes Codex CLI the better tool in general. It makes it the better tool when auditability and a no-cost start matter most.++{% img "note-2" %}++## Where does Claude Code actually win?++Two places: one engine across the tools you already use, and hooks. The terminal, VS Code, JetBrains, the desktop+app and the web all run the same Claude Code engine, so a repo's `CLAUDE.md`, settings and MCP servers behave the same+wherever you open it ([code.claude.com/docs/en/overview](https://code.claude.com/docs/en/overview), checked 10 Sep+2026). Hooks run shell commands before or after Claude Code's own actions, like formatting after every file edit or+running lint before a commit, so a rule you write once holds for every session instead of living in a README nobody+rereads. If you already pay for Claude, that is a lot of control for no extra setup.++## Can you run Codex CLI and Claude Code together?++Yes, and running both inside one harness is often more useful than picking a single winner. Munder Difflin wraps CLIs like `claude` and `codex` as real processes in their own terminals and runs them side by side on the same office floor, so a Codex agent and a Claude Code agent can each work on their own task at the same time. Munder Difflin 0.5.0 (2026-09-08) added GPT-6 Astra to Codex's model picker, and that slug needs Codex 0.153.1 or newer to be accepted when the agent spawns; the Codex CLI checked for this piece, `codex-cli 0.153.4`, clears that bar.++For the wider field beyond these two, the [2026 field guide to AI coding agents](/blog/best-ai-coding-agents/) covers where in-editor and autonomous agents fit next to CLI ones. For this specific pair, the honest answer holds: reach for Claude Code when you already pay for Claude and want hooks and one permission mode, reach for Codex CLI when you want separate sandbox and approval dials and a free way to start, and don't feel obligated to pick just one.diff --git a/blog/src/posts/how-much-does-claude-code-cost.md b/blog/src/posts/how-much-does-claude-code-cost.mdnew file mode 100644index 00000000..3d013b74--- /dev/null+++ b/blog/src/posts/how-much-does-claude-code-cost.md@@ -0,0 +1,67 @@+---+title: "How Much Does Claude Code Cost?"+description: "What Claude Code actually costs: free plan access, Claude Pro and Max pricing, per token API rates, and what running several agents does to the bill."+date: 2026-09-10+category: concepts+categoryLabel: Concepts+type: Non-technical+primaryKeyword: "how much does claude code cost"+secondaryKeywords: ["claude code pricing", "claude code cost per month", "is claude code free", "claude pro vs max", "claude code api pricing", "claude code cost india"]+tags: ["Concepts", "Cost", "Claude Code", "Getting Started"]+faq:+ - q: "Is Claude Code free to use?"+ a: "No, not on its own. Anthropic's free Claude plan lists chat, web search, memory and several other surfaces, but Claude Code is not among them; it first appears on Claude Pro and up (claude.com/pricing, checked 10 Sep 2026). The other way in is an Anthropic API key billed per token."+ - q: "How much does Claude Code cost per month?"+ a: "If you already pay for Claude Pro or Claude Max, nothing extra. Claude Pro is $20 a month billed monthly, or works out to $17 a month billed annually, and includes Claude Code at no added charge; Claude Max starts at $100 a month for more usage (claude.com/pricing, checked 10 Sep 2026)."+ - q: "What is the difference between Claude Pro and Claude Max for Claude Code?"+ a: "Mostly headroom. Anthropic describes the $100 a month Max tier as five times the usage of Pro per session, and a $200 a month tier as twenty times, both on a rolling five hour window with a weekly cap shared across every model (support.claude.com, checked 10 Sep 2026). Neither page states an exact Claude Code prompt count for any tier."+ - q: "How much does the Claude API cost per token for Claude Code?"+ a: "It depends on the model. As of 10 Sep 2026, Anthropic's pricing page lists Claude Sonnet 5 (Claude Code's default on Pro and Team Standard) at $2 per million input tokens and $10 per million output tokens, and Claude Opus 5 (the default on Max, Team Premium, Enterprise and a bare API key) at $5 and $25; Claude Haiku 4.5 runs $1 and $5 and is not a default anywhere (platform.claude.com/docs/en/about-claude/pricing and code.claude.com/docs/en/model-config)."+ - q: "Does Claude Code cost extra for teams or enterprise use?"+ a: "Team plans add a per seat charge on top of usage: a Standard seat is $20 a seat per month billed annually, a Premium seat is $100 a seat per month billed annually, and Enterprise starts at $20 a seat with usage billed on top (claude.com/pricing, checked 10 Sep 2026). All three still include Claude Code."+ - q: "How do I see what Claude Code is actually spending?"+ a: "Run /usage inside a session for a live breakdown by model and by attribution (skills, subagents, MCP servers), or the Session block for a running total in dollars computed at list price (code.claude.com/docs/en/costs, checked 10 Sep 2026). For an API key or cloud billed account, the Claude Console's usage page is the source of truth, not the CLI's own estimate."+---++Claude Code has no price of its own. As of 10 Sep 2026 it comes with Claude Pro at $20 a month and with Claude Max from $100 a month, the free Claude plan does not include it, and without a subscription you pay per token through the API (Anthropic, claude.com/pricing).++Which of those fits depends on how you already work. If you are the kind of developer running more than one Claude Code window already, it is worth reading [how to keep several Claude Code sessions straight](/blog/manage-multiple-claude-code-sessions/) first, because tracking spend gets harder exactly when you add more terminals. The rest of this post breaks down each option, names its source, and says when it was checked, because Anthropic changes these numbers without much warning.++## Is Claude Code free to use?++No. Anthropic's free Claude plan covers chat, web search, memory, file creation, code execution and a few other surfaces, but Claude Code does not appear on that list; it starts on Claude Pro (claude.com/pricing, checked 10 Sep 2026). The other route in skips the subscription entirely: connect an Anthropic API key and Claude Code bills per token instead.++## How much does Claude Code cost with a Claude subscription?++Claude Pro is $20 a month billed monthly, or $17 a month on the annual plan billed up front, and Claude Code comes included at no extra charge (Anthropic, claude.com/pricing, checked 10 Sep 2026). Claude Max sits above it at $100 a month for a five times usage tier and $200 a month for a twenty times tier, with both figures set against Pro's own usage, and both plans meter Claude Code and Claude chat together rather than as two separate allowances ([support.claude.com](https://support.claude.com/en/articles/11145838-use-claude-code-with-your-pro-or-max-plan), checked 10 Sep 2026).++{% img "note-1" %}++## What's the difference between Claude Pro and Claude Max for Claude Code?++Mostly how much you can burn through before you wait. Anthropic states usage on a rolling five hour window plus a weekly cap shared across models, and Max simply multiplies that ceiling by five or twenty depending on which tier you pick ([support.claude.com](https://support.claude.com/en/articles/11049741-what-is-the-max-plan), checked 10 Sep 2026). What neither official page states is an exact number of Claude Code prompts per window; specific figures like "10 to 40 prompts" that circulate on comparison blogs are not on Anthropic's own pages as checked for this post, so treat them as someone's estimate rather than a published limit.++## How much does the Claude API cost per token?++It scales with the model. Which model Claude Code even starts on depends on your plan: Sonnet 5 by default on Pro and Team Standard, Opus 5 by default on Max, Team Premium, Enterprise and a bare API key (code.claude.com/docs/en/model-config, checked 10 Sep 2026). As of 10 Sep 2026, Anthropic's pricing page lists Claude Sonnet 5 at $2 per million input tokens and $10 per million output tokens, Claude Opus 5 at $5 and $25, and Claude Haiku 4.5 at $1 and $5 (platform.claude.com/docs/en/about-claude/pricing). Routine edits on a cheap model and hard reasoning on an expensive one, the habit our guide on [matching tasks to the right model size](/blog/do-more-with-less-model-routing/) covers, is where most of the saving actually lives. The same page notes that Sonnet 5's $2/$10 rate, originally an introductory price, became the standard rate once August 2026 ended, so it is not a limited time deal anymore. Prompt caching cuts the input side further, down to roughly a tenth of the base rate on a cache hit, the same lever our [multi-agent cost playbook](/blog/the-multi-agent-cost-playbook/) covers for a whole fleet of agents rather than one session.++{% img "note-2" %}++## Does Claude Code cost more in India?++Anthropic does not publish a separate India price on the pages checked for this post. Claude.com/pricing showed only USD figures as fetched on 10 Sep 2026 for this check, and neither the Pro nor the Max support article names an INR amount. Rupee figures circulating on third party blogs are not traceable to an Anthropic page, so this post is not repeating them; check claude.com/pricing directly for the number your account sees.++## What does running several Claude Code agents do to the bill?++It multiplies, because every agent is its own Claude Code process with its own context window and its own token meter running in parallel, not a shared allowance split between them. Munder Difflin, the harness this blog belongs to, does not change what any of that costs: every agent it hires still runs on your own Claude Code subscription or API key, never a resold or marked up rate. What it adds is a lever on the multiplication rather than a discount on it: per agent token budgets cap what a single hired agent can spend, and a circuit breaker steers, then constrains, then stops an agent that loops, storms errors, or blows its budget, instead of a person finding out from the invoice ([Munder Difflin README](https://github.com/chaitanyagiri/munder-difflin#readme), checked 10 Sep 2026).++Claude Code itself carries the same instinct for a single run. Checked on 10 Sep 2026, running `claude --help` on Claude Code 2.1.267 turns up its own per session guard:++```+--max-budget-usd <amount> Maximum dollar amount to spend on API+ calls (only works with --print)+```++Set that on one agent and it stops on its own. Set it on ten, running at once with nobody watching, and it is the difference between a surprise and a line item; Oscar in accounting would call that the bare minimum. Claude Code's own `/usage` command and Session block give a running dollar estimate at list price for a single session too, covered in Anthropic's [cost management docs](https://code.claude.com/docs/en/costs), though a fleet of separate agents means reading that number once per agent rather than once for the whole job.++None of the figures above are permanent. Anthropic revises pricing on its own pages with little notice, the way Sonnet 5's introductory rate became its standard rate after 31 August 2026, so treat this post as accurate for 10 Sep 2026 and check [claude.com/pricing](https://claude.com/pricing) or the [Claude Platform pricing docs](https://platform.claude.com/docs/en/about-claude/pricing) before you commit a team to a plan.diff --git a/blog/src/posts/how-to-add-an-mcp-server-to-claude-code.md b/blog/src/posts/how-to-add-an-mcp-server-to-claude-code.mdnew file mode 100644index 00000000..5c81636e--- /dev/null+++ b/blog/src/posts/how-to-add-an-mcp-server-to-claude-code.md@@ -0,0 +1,114 @@+---+title: "How to Add an MCP Server to Claude Code"+description: "The claude mcp add command, its scope flag, editing .mcp.json by hand, and how Munder Difflin's Capabilities screen grants an MCP server to one agent at a time."+date: 2026-09-10+category: guides+categoryLabel: Guides+type: Technical+primaryKeyword: "how to add an mcp server to claude code"+secondaryKeywords: ["claude code mcp add", "claude code mcp json", "how to install mcp server claude code", "add mcp server to claude code globally", "claude code mcp add scope"]+tags: ["Guides", "MCP", "Claude Code", "Getting Started"]+faq:+ - q: "What is an MCP server?"+ a: "A program that gives an agent tools or data it does not have built in, such as a database, an issue tracker or a browser, and exposes them over the Model Context Protocol. Claude Code is one of many MCP clients, so once a server is added, its tools sit alongside Claude Code's built in ones."+ - q: "How do I add an MCP server to Claude Code from the command line?"+ a: "Run claude mcp add with a name you choose, then the command that starts the server after the separator Claude Code expects, or, for a hosted server, add a URL with the transport flag set to http. Run claude mcp list afterward to confirm it connected."+ - q: "What is the difference between local, project and user scope?"+ a: "Local scope, the default, is private to you and only active in the project where you added the server. Project scope writes the entry to a .mcp.json file you commit, so teammates who clone the project get the same server. User scope registers the server once for every project you open, still private to you."+ - q: "How do I add an MCP server that needs an API key?"+ a: "Pass it with the -e flag when you add the server, for example -e API_KEY=xxx. A server that signs in through a browser instead, such as ones that use OAuth, is added with just its URL and authenticated afterward from inside a session with the /mcp command."+ - q: "Does Munder Difflin let me control which agent gets which MCP server?"+ a: "Yes. Since Munder Difflin 0.4.7, the Capabilities screen in the Pro workspace can turn a server on for one agent and off for another, and servers that can write or hold secrets still need an explicit yes. Each Claude Code agent still picks up the servers you added at user or project scope, the same as any other claude session."+---++<div class="callout tldr"><span class="ic">TL;DR</span><p>Run <code>claude mcp add</code> with a name and the command that starts the server for a local tool, or add the transport flag set to <code>http</code> and a URL for a hosted one. The default scope is local to your project; add the scope flag set to <code>user</code> to use a server everywhere, or to <code>project</code> to share it with teammates through <code>.mcp.json</code>. Confirm the connection with <code>claude mcp list</code>, and edit <code>.mcp.json</code> directly if you would rather write the JSON yourself.</p></div>++You add an MCP server to Claude Code with one command: `claude mcp add`, a name you pick, and the command or URL that starts the server. Claude Code writes that entry to a config file, then loads the server's tools the next time it reads it.++That one command handles the common case. What's worth understanding is the scope flag, which decides who else sees the server, and the other ways to add one: editing `.mcp.json` yourself, or connecting from VS Code or the desktop app instead of the CLI. If you're already running more than one Claude Code session, each reads the same project scoped servers, which is one less thing to juggle when you [manage multiple Claude Code sessions](/blog/manage-multiple-claude-code-sessions/).++## What is an MCP server?++An MCP server is a program that gives an agent tools or data it doesn't have built in: a way to search issues, query a database, or drive a browser. The Model Context Protocol is the open standard those servers speak, described on its own site as "a standardized way to connect AI applications to external systems" (checked 10 Sep 2026, [modelcontextprotocol.io](https://modelcontextprotocol.io)). Anthropic open sourced MCP on 25 November 2024, with initial SDKs and a handful of reference servers (checked 10 Sep 2026, [Introducing the Model Context Protocol](https://www.anthropic.com/news/model-context-protocol)). Claude Code is one client among several now; any compliant server works with it.++## How do you add an MCP server to Claude Code?++Run `claude mcp add`, pick a name, and tell it what starts the server: a local command, or a URL with the transport flag set to `http` for a hosted one. Here's what that looked like in a scratch folder, checked on 10 Sep 2026:++```bash+$ claude --version+2.1.267 (Claude Code)++$ mkdir -p mcp-demo && cd mcp-demo+$ claude mcp add --scope project demo-docs -- npx -y @modelcontextprotocol/server-everything+Added stdio MCP server demo-docs with command: npx -y @modelcontextprotocol/server-everything to project config+File modified: <your folder>/mcp-demo/.mcp.json++$ cat .mcp.json+{+ "mcpServers": {+ "demo-docs": {+ "type": "stdio",+ "command": "npx",+ "args": ["-y", "@modelcontextprotocol/server-everything"],+ "env": {}+ }+ }+}+```++`claude mcp add` wrote that block the moment the command ran; nothing else to save or restart beyond starting a fresh session in the folder. For a hosted server you swap the local command for the transport flag set to `http` and a URL, the same shape Anthropic's own quickstart uses for its docs server (checked 10 Sep 2026, [Connect to MCP servers](https://code.claude.com/docs/en/mcp-quickstart)).++{% img "note-1" %}++## What's the difference between local, project and user scope?++Local scope is the default: private to you, active only in the project where you ran the command. Project scope writes to the `.mcp.json` file you just saw above, which you commit so teammates get the same server; user scope registers a server once for every project you open, still private to you. Michael Scott would put everything in project scope and commit his API key for the whole office to read.++Pick project scope for anything the team needs in the repo, like a shared database or issue tracker server. Pick user scope for tools you personally want everywhere, like a search server you reach for across projects. Leave local scope for anything you're just trying out. The demo above used project scope on purpose, so you could see the `.mcp.json` file it writes.++## How do you add a server that needs an API key?++Pass it with the `-e` flag when you add the server, for example `-e API_KEY=xxx` after the server's name and command. A server that authenticates through a browser instead, like the hosted examples in Anthropic's own docs, is added with just its URL, then signed in afterward from inside a session with the `/mcp` command.++{% img "note-2" %}++## Can you edit .mcp.json directly instead of using the CLI?++Yes. Every scope writes the same JSON shape, so you can hand write an entry instead of running the command, which helps when a server needs several environment variables you'd rather template in one file:++```json+{+ "mcpServers": {+ "my-server": {+ "type": "stdio",+ "command": "npx",+ "args": ["-y", "<the server's npm package>"],+ "env": {+ "API_KEY": "your key here"+ }+ }+ }+}+```++Claude Code reads `.mcp.json` at the start of a session, not while one is running. And a project scoped server needs a yes the first time Claude Code sees it: when we ran `claude mcp list` against the demo folder above, it showed `demo-docs` as pending approval rather than connected, which is the point. Approve a server because you've read what it does, not on reflex; a server's own tool descriptions are text Claude reads, and a rogue change to them is exactly the kind of attack [this blog has covered before](/blog/mcp-security-tool-poisoning/).++## Do MCP servers work the same way in VS Code or the desktop app?++Mostly, with a different entry point. VS Code and the Claude Code desktop app each have their own place to add a server instead of a terminal command, and Claude Code on the web just reads whatever `.mcp.json` is already checked into the repository. The CLI steps above are for a terminal session. The desktop app and the CLI read the same `~/.claude.json` and `.mcp.json` files, although a server in the desktop app's own `claude_desktop_config.json` wins when the names collide (checked 10 Sep 2026, [Claude Code desktop docs](https://code.claude.com/docs/en/desktop)).++## Can Munder Difflin grant an MCP server to just one agent?++Yes. Since Munder Difflin 0.4.7, the Capabilities screen in the Pro workspace can turn a server on for one agent and+off for another on the same floor, and servers that can write or hold secrets still need an explicit yes.++That sits on top of Claude Code's own setup, not instead of it. Each Claude Code agent in the office is a real+`claude` session, so it still picks up the servers you added at user or project scope, as+[an earlier post on MCP in a hive](/blog/mcp-and-skills-in-a-hive/) explains. Capabilities is the extra dial for+keeping a server away from one agent that could otherwise reach it.++None of this needs Munder Difflin: `claude mcp add` is Claude Code's own command and works the same in a bare+terminal. A harness earns its place once several agents are running and only one of them should hold a given key.+[Download Munder Difflin](https://munderdiffl.in/download); the classic office is free, and Capabilities comes with+[Pro](https://munderdiffl.in/#pricing).diff --git a/blog/src/posts/mcp-and-skills-in-a-hive.md b/blog/src/posts/mcp-and-skills-in-a-hive.mdindex e6529289..efa0697a 100644--- a/blog/src/posts/mcp-and-skills-in-a-hive.md+++ b/blog/src/posts/mcp-and-skills-in-a-hive.md@@ -57,7 +57,7 @@ agent inside it. A harness wraps the agent you already run — which is the coreMCP (Model Context Protocol) servers extend an agent with external tools and data — your database, yourissue tracker, your internal APIs. Claude Code resolves MCP servers from configuration: user-scoped-servers you've set up globally, and project-scoped servers defined in the project you're working in.+servers you've set up globally, and project-scoped servers defined in the project you're working in. If you haven't set one up yet, [our guide to adding an MCP server to Claude Code](/blog/how-to-add-an-mcp-server-to-claude-code/) walks through the `claude mcp add` command and its scope flags.Because a hive agent runs as a real session in that project, it picks up both. Every agent in the hivecan call the same MCP tools you'd have in a solo session — no per-agent re-registration.diff --git a/blog/src/posts/the-multi-agent-cost-playbook.md b/blog/src/posts/the-multi-agent-cost-playbook.mdindex 35e12052..20794e6e 100644--- a/blog/src/posts/the-multi-agent-cost-playbook.md+++ b/blog/src/posts/the-multi-agent-cost-playbook.md@@ -79,6 +79,8 @@ This is a concrete feature in Munder Difflin, not just advice: the harness's peragents and an Opus-class model to the GOD orchestrator **today** — the cheap-workers/premium-lead patternconfigured directly.+For what each Claude plan and model costs as of September 2026, see [how much Claude Code costs](/blog/how-much-does-claude-code-cost/).+## Lever 4 — Context discipline: don't pay to re-read everything, N timesThe multiplier nobody budgets for: every agent's context window is a **recurring per-turn cost**, and in