BlogWhat the numbers say.
Every guide, comparison, price list and finding, newest first. Measured on a real machine, dated.
Comparisons · updated
Gemini CLI vs Claude Code: the end of the free tier, Google plans, limits and cost
Gemini CLI vs Claude Code after the free tier ended: plans, limits, cost, and a pointer to Kiro CLI, AWS's agent that replaced the Amazon Q Developer CLI.
Pricing · updated
Claude Code on Bedrock: cost, and who sees it
Run Claude Code through Amazon Bedrock, see where the spend lands on the AWS bill, why Anthropic's analytics never see it, and how to tie it to a repository.
Guides · updated
Claude Code OpenTelemetry: the variables, the metrics, and what the stream never carries
The variables that export Claude Code metrics to your collector, which attributes name a person, the one repository field, and what the stream never carries.
Guides · updated
Claude Code privacy: what leaves the laptop
What Claude Code sends to Anthropic on each plan, what telemetry and analytics add about a person, and how to measure agent spend by repository, not by person.
Guides · updated
Claude Code stuck: hung, looping, or waiting
Tell a hung Claude Code session from a loop or a waiting prompt, what to do for each, and how a team sees one repeating a call five times in three minutes.
Pricing · updated
Claude Team plan: which seat runs Claude Code
Standard or Premium seat for Claude Code, what each includes, where the limit sits, and how measured sessions tell whether seats or the API cost less.
Guides · updated
Codex analytics for a team, by repository
What OpenAI shows an admin about Codex usage, what the usage dashboard shows one person, and how to see Codex and Claude Code in one table by repository.
Guides · updated
FinOps for coding agents: allocate, report, cap
Allocate Claude Code, Codex and Bedrock agent spend to repositories and pull requests, report it to finance beside the invoice, and cap the loop, not the work.
Pricing · updated
Claude Code pricing: every plan, the API rates, and what real use costs
Claude Code pricing on every plan — Pro, Max, Team, Enterprise — the API rates per token, and what real use costs, measured on one machine.
Pricing · updated
Claude Enterprise pricing: seats plus usage
What Anthropic publishes about Enterprise seats and usage billing for Claude Code, what it does not, and how to size a plan from measured cost per PR.
Guides · updated
Claude Code memory: CLAUDE.md, auto memory, /memory, and what it still forgets
How Claude Code memory works: CLAUDE.md, auto memory and MEMORY.md, the /memory command, its limits, and how to keep and search every past session.
Comparisons · updated
cmux on Windows and Linux: what exists today, and the alternatives for running coding agents
cmux is a macOS app; Windows is a waitlist, Linux gets another product. On Windows and Linux: Herdr, a cockpit in your terminal, or Caprock, the ledger.
Comparisons · updated
Desktop apps for coding agents compared: Caprock, Orca, Conductor, Claude, Codex, Warp and Nimbalyst
Seven apps that run Claude Code, Codex and other coding agents side by side: licence, desktop OS, agents, how a phone reaches you, cost and limits, price.
Comparisons · updated
Herdr vs cmux: a terminal runtime or a Mac terminal app for coding agents, and where Caprock fits
Herdr is a cockpit in your own terminal on macOS, Linux and Windows; cmux is a macOS terminal app with paid cloud and seats. Phone, price and what each sees.
Pricing · updated
Is Codex free? Codex pricing, plans and usage, and what a session costs
Is Codex free? Codex pricing on every ChatGPT plan and the API, how to check your usage on the dashboard or with /status, and what a session costs.
Comparisons · updated
Orca IDE vs Caprock: a cockpit for coding agents, and the ledger of what they did
Orca IDE is a cockpit that starts Claude Code, Codex and more in worktrees; Caprock is the ledger of every session on the machine. Read 27 October 2026.
Comparisons · updated
Orca vs Herdr: a desktop app or a terminal runtime for coding agents, and where Caprock fits
Orca is a desktop cockpit for coding agents; Herdr is one inside your own terminal. SSH, phone, price and what each sees, from each maker's own pages.
Comparisons · updated
T3 Code vs Caprock: two open-source apps for running coding agents, compared
T3 Code is a chat cockpit over a server you can host, with phone apps; Caprock is the ledger of every agent session on the machine. Read 27 October 2026.
Guides · updated
Claude Code terminal: how to open it, the best terminal for it, and the setup that matters
Claude Code in a terminal: how to start it, which terminal handles Shift+Enter and notifications, tmux and Windows setup, and Caprock's app as a terminal.
Guides · updated
How to check Claude Code usage: tokens, cost and plan limits
Check Claude Code usage with /usage for tokens, cost and plan limits. Use Caprock, a free desktop app, to track past sessions and costs by repository.
Guides · updated
Claude Code context window: 1M or 200K, auto-compact, /compact and /clear
Claude Code's context window per model, how to see usage, when auto-compact fires and how to change or turn it off, /compact vs /clear, and what context costs.
Guides · updated
Claude Code MCP: how to add a server, which scope to use, and examples that work
Add an MCP server to Claude Code with claude mcp add: HTTP and stdio, local, project and user scope, .mcp.json, OAuth, limits, and working examples.
Guides · updated
Claude Code mobile: use it on your phone with Remote Control, or with Caprock from anywhere
Claude Code on mobile: the Claude app for iOS and Android with Remote Control, or Caprock to start work, chat, commit and push from anywhere, no account.
Guides · updated
Claude Code notifications: get told when it finishes or needs you, on your desktop, phone or Telegram
Get a Claude Code notification when it finishes or waits for approval: the built-in alert, Notification hooks, phone push, Telegram, and Caprock's free alerts.
Comparisons · updated
Codex vs Claude Code: pricing and usage limits at $20, $100 and $200, and cost measured on one machine
Codex vs Claude Code pricing and usage limits, plan by plan: ChatGPT Plus vs Claude Pro at $20, Pro vs Max at $100 and $200, and cost per call measured.
Comparisons · updated
AI coding agents in 2026: Claude Code, Codex, Cursor, OpenCode and Gemini CLI compared
AI coding agents compared: Claude Code, Codex, Cursor, OpenCode and Gemini CLI on price, limits, models and licence, with cost per call measured.
Guides · updated
claude --dangerously-skip-permissions: what it does, the risks, and safer options
What claude --dangerously-skip-permissions turns off, what still asks, how to run it safely in a container or sandbox, and the safer options: auto mode, rules.
Guides · updated
Claude Code desktop app: the official app, desktop vs CLI, and a free app that runs your coding agents
The Claude Code desktop app: Anthropic's Code tab, what it adds over the CLI, the web and IDE versions, and Caprock, a free app that records every agent.
Guides · updated
Claude Code GUI: the official apps, open-source GUIs like opcode, and a free desktop app for every agent
Claude Code GUI options: Anthropic's desktop app, VS Code and web UI, open-source GUIs (opcode, CloudCLI, Nimbalyst), and Caprock, a free app for every agent.
Guides · updated
Claude Code history: where conversations are stored, and how to keep them
Where Claude Code stores conversation history, why it disappears after 30 days, how to resume or export a session, and how to keep and search all of it.
Guides · updated
Claude Code models: what each alias means, how to change model, and what each costs
Which models Claude Code can use, what opus, sonnet, haiku, fable and opusplan mean today, how to change model, and what each cost on one machine.
Guides · updated
Claude Code multiple agents: how to run them in parallel, and check the work
Run several Claude Code agents in parallel with subagents, agent teams, worktrees or -p, then keep one record of them and check the work with Caprock.
Guides · updated
Claude Code Remote Control: how to use it, and why it fails to connect
Claude Code Remote Control: three ways to start it, what it needs, each error and its fix, Dispatch compared, and Caprock for the sessions it cannot reach.
Guides · updated
Claude Code status line: how to set it up, and what to show
Set up Claude Code's status line in settings.json or with /statusline, what data it receives, what it cannot show, and how Caprock fills the gap.
Guides · updated
Claude Code usage monitors: 5 tools compared, and what each is best for
Five Claude Code usage monitors and trackers compared: ccusage, Claude-Code-Usage-Monitor, CodexBar, Caprock, tokscale. What each shows, and its limits.
Guides · updated
Claude Code worktrees: run parallel sessions with --worktree, set them up, and clean them up
Claude Code worktrees: start a session in its own checkout with --worktree, carry .env over, clean up, and see every worktree's sessions and diff in Caprock.
Pricing · updated
Claude usage limits: the 5-hour session limit, the weekly limit, and when they reset
Claude and Claude Code usage limits on Free, Pro, Max and Team: the 5-hour session limit, the weekly limit, when each resets, and how close you are.
Guides · updated
Codex desktop app: OpenAI's app on Mac, Windows and Linux, and Codex CLI in tabs with its cost
The Codex desktop app is now the ChatGPT desktop app: how to get it on Mac, Windows or Linux, what it does, and Codex CLI in Caprock's tabs with its cost.
Guides · updated
Codex remote control: from your phone, on Mac, Windows or Linux
Codex remote control: OpenAI's phone pairing for Mac and Windows, why it fails on Linux and some Windows PCs, and Codex from your phone on any OS with Caprock.
Pricing · updated
Cursor pricing: Plus, Premium, Super, Ultra and Teams, and what Claude costs inside
Cursor pricing: Free, Plus at $8, Premium $30, Super $100, Ultra $200 and Teams; the weekly pool, Claude billed on top, and what became of Pro and Pro+.
Comparisons · updated
Cursor vs Claude Code pricing: each plan, usage limits, and what Claude costs in Cursor
Cursor vs Claude Code pricing and usage limits at $20, $100 and $200: no Cursor plan includes Claude, which bills at API price; Claude plans are a flat fee.
Comparisons · updated
Cursor vs Codex: plans, prices and measured usage
Cursor vs Codex on plans, prices and usage limits, what GPT models cost in each, and what Codex calls cost, measured on one machine that runs it.
Comparisons · updated
Google Antigravity vs Claude Code: free quota, Google AI plans and cost
Google Antigravity vs Claude Code on price, the free quota, Google AI Pro and Ultra limits, the Claude models Antigravity runs, and what Claude Code costs.
Guides · updated
How to install Claude Code on Mac, Windows and Linux
Install Claude Code with one command on macOS, Windows or Linux: native installer, Homebrew, WinGet or npm, how to sign in, VS Code and JetBrains, and fixes.
Guides · updated
How to install Codex CLI on Mac, Windows and Linux
Install OpenAI's Codex CLI with one command on macOS, Windows or Linux: the installer, npm or Homebrew, native Windows or WSL, how to sign in, and fixes.
Guides · updated
How to install OpenCode on Mac, Windows and Linux
Install OpenCode on macOS, Windows or Linux: the script, npm, Homebrew, Scoop or pacman, OpenCode 1 or 2, first run, updates, fixes, and OpenCode in Caprock.
Guides · updated
How to resume a Claude Code session, and find the one you closed
Resume a Claude Code session with --continue, --resume or /resume, list past sessions in the picker, fix a failed resume, and continue one from Caprock.
Guides · updated
How to resume a Codex session, and find old ones
Resume a Codex session with codex resume, --last or an ID, find old ones in ~/.codex/sessions, desktop app threads included, and continue them in Caprock.
Pricing · updated
Is Claude Code free? No — here is the cheapest way in, and the agents that start at $0
Is Claude Code free? No: Claude Free leaves it out. The cheapest way in is Pro at $20 or an API key, and Antigravity, OpenCode and Codex start free.
Comparisons · updated
Open-source AI coding agents: which are open, where they run, and what they cost per call
Open-source AI coding agents: OpenCode, Codex CLI, Gemini CLI, Cline, goose and more by licence and form, with cost per call measured on one machine.
Guides · updated
OpenCode desktop app: the official beta on Mac, Windows and Linux, and OpenCode in tabs with its cost
OpenCode Desktop is a beta app for macOS, Windows and Linux: how to install it, how it differs from the TUI, and OpenCode in Caprock's tabs with its cost.
Pricing · updated
OpenCode pricing: Go, Zen and your own key, and what a call really costs
OpenCode is free; the model is not. OpenCode Go at $10 or $40 a month, Zen per token, or your own key — with each price, and what a call cost on one machine.
Comparisons · updated
OpenCode vs Claude Code: models, pricing and cost, measured on one machine
OpenCode vs Claude Code on models, pricing and Claude subscriptions, and what each cost per call on one machine that runs both. Not a verdict on code quality.
Guides · updated
How to install Claude Code on Linux: Ubuntu, Fedora, Arch, WSL, and a desktop app
Install Claude Code on Linux with the native installer, apt, dnf, the AUR or npm; fix PATH and WSL errors, update, uninstall, and run it in a desktop app.
Guides · updated
Claude Code hooks: how they work, with examples you can copy
Claude Code hooks run your commands at points in a session. The events, the settings.json format, exit codes, and six tested examples to copy into your project.
Guides · updated
Codex usage limits: the 5-hour and weekly windows, by plan
Codex usage limits on Plus, Pro and Business: the 5-hour window and the weekly limit, how to check and reset them, and what one machine's windows recorded.
Comparisons · updated
OpenCode vs Codex: plans, limits and both measured on one machine
OpenCode vs Codex: price, plans, usage limits and models, and what each agent's calls cost, both read by Caprock on one machine that runs the two.
Guides · updated
Claude Code subagents: how to create them, when Claude uses them, and what they cost
Claude Code subagents: the built-in ones, writing your own in .claude/agents/, forcing one with @agent, background runs, and what they cost on one machine.
Pricing · updated
GitHub Actions pricing: per-minute rates, free minutes and a cost calculator
GitHub Actions pricing as of 12 October 2026: every runner's per-minute rate, free minutes by plan, storage, and a calculator that takes the free minutes off.
Comparisons · updated
GitHub Copilot vs Claude Code: pricing, AI credits and what heavy use costs
GitHub Copilot vs Claude Code on plans, Copilot's AI credits since June 2026, what Claude costs inside Copilot, and what heavy Claude Code use costs.
Comparisons · updated
Kiro vs Claude Code: credits, plans, Claude in Kiro, and spec-driven work
Kiro vs Claude Code on plans and credits, what each Claude model costs in Kiro, spec-driven work against plan mode, and what heavy Claude Code use costs.
Guides · updated
Claude Code auto mode: what it allows, what still asks, and how to switch it
Claude Code auto mode: turn it on or off, what the classifier blocks, why it still asks, auto vs bypass, and the prompts that remain answered in one place.
Guides · updated
Claude Code system prompt: what is in it, and how to append to it or replace it
What Claude Code's system prompt holds, how to add to it with --append-system-prompt or replace it, where CLAUDE.md and output styles fit, and per-project use.
Guides · updated
CLAUDE.md: what to put in it, where it goes, and examples
What CLAUDE.md is, where Claude Code looks for it, what to put in it and what to leave out, with an example file, imports, rules, AGENTS.md and auto memory.
Guides · updated
Claude Code API key: how to use one, and when it costs less than a plan
Get a Claude Code API key, make Claude Code use it or switch back to your plan, and see with Caprock whether the API or a plan costs you less.
Guides · updated
Claude Code skills: how to add and install them, and how Claude picks one
Claude Code skills: what they are, how to add one with SKILL.md or install one from a marketplace, how Claude picks one, and what one machine invoked.
Research · updated
49% of my Claude Code tool calls ran commands
In 58 days of my Claude Code sessions, 49% of tool calls ran shell commands. The call counts, cost at API list prices and cache figures from one machine.
Pricing · updated
Claude API pricing: every model, and what one call costs
Claude API pricing per million tokens for Opus, Sonnet, Haiku and Fable, with cache and batch rates, a calculator and 79,014 real calls priced.
Research · updated
Claude Code analytics: what Anthropic shows a team
Anthropic gives a team a dashboard, a spend report, two APIs and OpenTelemetry. What each one shows, what it does well, and what none of them answer.
Research · updated
Claude Code cost per month: $5,898 on a $200 plan
My August of Claude Code on a $200 Max plan came to $5,898 at API list prices — 29 times the fee. Three months of usage from one machine.
Research · updated
Claude Code had already deleted half my history
It keeps transcripts for 30 days. On my machine that was 67 sessions and $7,808 of work — gone from disk, still searchable in Caprock.
Research · updated
Five coding agents, five formats, one table
Claude Code, Codex, OpenCode, Gemini CLI and DeepSeek Harness each record their work differently, and three of the differences give a believable wrong number.
See every coding agent you run.
Who is waiting on you, who is stuck, how close the limit is. Free and open source; nothing leaves your machine.
Notarized by Apple. One file for Apple silicon and Intel, macOS 13 or later.
Or run it yourself:
Paste in Terminal: it installs Caprock and opens it.Or with Homebrew: brew install --cask dspv/tap/caprock-app