A coding agent harness: one session, on any machine, model or agent.
Krowk is an agent harness built for power users who run coding agents all day. It runs Claude, GPT, Grok or any OpenAI-compatible model, using your existing subscription or an API key, or drives Claude Code and Codex. A session can move to another model or agent mid-conversation, and syncs to your other machines end-to-end encrypted. It follows Claude Code's permissions, skills and hooks, keeps track of every agent session on your machine, and publishes screenshots, diffs and logs as links that unfurl in GitHub, Slack, Linear and Basecamp. It assumes you're comfortable in a terminal, with git and with configuration files.
Publishing from Krowk, the coding agent harness, for any MCP client: an MCP server that turns an agent's screenshots, recordings, diffs and logs into links that unfurl in Slack and Basecamp; in GitHub and Linear, images show inline.
2026-11-11
ดาว
—
เครื่องมือ
—
อัปไทม์
—
krowk connect # sign in with a Claude, ChatGPT or SuperGrok subscription, or an API key
krowk # open the agent
krowk -p "fix the build" # run one prompt headless (--output-format json)
krowk --resume # continue an earlier session
Builds are published for Linux and macOS (amd64/arm64) and Windows (amd64). On a workstation the installer gives you the full build, which includes the agent. In CI and containers it gives you the lean build, which only publishes. Pass --full or --lean to choose (curl … | bash -s -- --full).
The agent
A prompt that stays put. The agent takes the whole terminal, and the prompt and status line stay on the bottom rows while you scroll the conversation with the mouse wheel or PgUp/PgDn, and dragging over it copies. When you leave, your shell comes back as it was, with the krowk --resume <id> line to open the session again. Choose Screen: inline in /settings (or set "tui": {"screen": "inline"} in ~/.krowk/config.json) to draw at the bottom of the terminal instead, with everything in your normal scrollback as it streams, which suits SSH, tmux and phone terminals.
Tools. It can read, write, edit, run bash, grep, glob and publish. It also keeps a todo list, runs subagents in parallel, each with its own context and a cheaper model, and asks you with options to pick from when a choice is yours.
Compatible with Claude Code. It uses the same permission modes, allow/ask/deny rules and settings files. It reads AGENTS.md, CLAUDE.md and .cursor/rules, and loads Claude-format skills, hooks and .claude/agents without changes. It adds an unhinged mode that never asks for permission.
Multiple accounts. You can connect several accounts per provider and rename them in /connect. A second account is added with --name work.
Cost limits.--max-usd and --max-tokens stop a session before it goes over a limit, subagents included.
In the prompt, / lists the commands and ? shows the keys.
Worktrees
Agents running at once can each work in a git worktree of their own, on a branch krowk/<hex>, so they don't overwrite each other's files.
Subagents.isolation: "worktree" on the subagent call, or isolation: worktree in an agent definition's frontmatter, runs the child in a new worktree started from your files as they are, uncommitted changes included. When it finishes, an unchanged worktree is removed; a changed one is applied back to your working tree (never your index or HEAD) and removed. A conflict, a submodule change or a change to .git, .claude, .codex or .krowk keeps it, and the summary names it.
A session of its own.krowk --worktree (or krowk -p --worktree) starts the session in a new worktree. On exit an unchanged one is removed and a kept one is named.
By default."worktrees": { "default": true }, in your config or a trusted repository's .krowk/config.json, starts every new session in a repository that way, whatever model or agent runs it. A worktree it can't make stops the session rather than letting it run in your checkout; --worktree=false runs one in place. As with --worktree, the session runs in krowk's own process rather than the background host, so it ends with the terminal, and /new carries on in the same worktree.
No hand-made ones. Whatever the backend or mode, an agent's git worktree add, or Claude Code's EnterWorktree and Agent with isolation: "worktree", is refused and pointed to krowk's worktrees. So is an upload to a paste site (gh gist create, curl -F to 0x0.st, transfer.sh and the like), pointed to publish. You can still run any of these at your own prompt.
Ready to build. Submodules are initialised from your checkout, target and node_modules are cloned in where the file system makes that nearly free (worktrees.seed), the ignored files your .worktreeinclude lists (.env, local config) are copied, and worktrees.setup runs a command such as npm ci in the sandbox. A repository's own setup command runs only once you trust it. On btrfs a worktree is a snapshot, made in tens of milliseconds whatever the repository's size.
The machine's limits. Build and test commands (cargo, make, npm, pytest, …) from every krowk on the machine queue for builds.slots slots, a quarter of the cores by default. Subagents and --worktree sessions share subagents.maxHost agent slots. A worktree is refused when its disk is short of 4 GiB beyond twice its expected size.
Command
What it does
krowk worktrees
List krowk's worktrees: base, commits ahead, uncommitted changes, whose session
krowk worktrees apply <hex> [--to <dir>]
Bring a kept worktree's changes into your checkout, uncommitted
krowk worktrees remove <hex> [--force]
Remove one; --force saves its changes as a ref first, and the branch is kept
krowk worktrees prune
Clear what deleted worktrees left, and snapshots over 30 days old (also runs daily)
scripts/bench-worktrees measures all of this on this repository: 20 worktrees made and checked at once, and three subagents applied back.
Sessions
The full build keeps a local store of your agent sessions, whether they came from krowk, Claude Code, Cursor or opencode.
Command
What it does
krowk sessions
List every session on this machine, newest first
krowk sessions show <id>
Read a session back, turn by turn
krowk sessions import / sync
Import transcripts and reconcile usage
krowk sessions budget <id> --max-usd 5
Exit 4 when a session costs more than the limit, for use in a hook
Sync
Sessions sync between your machines end-to-end encrypted: neither krowk's servers nor its relay can read them. Run a session on one machine, then follow it and prompt it from another.
Once sync is set up, every session that runs a turn in the host daemon syncs by itself. Set "sync": {"auto": false} in ~/.krowk/config.json to turn that off, or keep one directory off sync with sync.exclude or a .krowk/no-sync file. A session nobody has used or watched for 10 minutes (sync.idleMinutes) stops syncing until its next prompt on this machine, or until another device, your phone included, opens it while the host daemon runs and has stayed online. krowk sync status says what applies where you are, and krowk host enable keeps sessions synced on an always-on machine.
Command
What it does
krowk sync init
Set up sync on this machine, with a recovery kit
krowk devices add / krowk sync join
Show a code on one machine, enter it on the new one
krowk sync status
Your device list, and whether sessions here sync by themselves
krowk sync host <session>
Sync one session by hand, an opted-out one included
✓ Uploaded screenshot.png → https://krowk.com/a/art_2e1d
412 KB · run run_8Kd2wq
Each link shows a card with the repo, commit, branch, PR, session and agent, all detected from git and CI. You can push without a key: anonymous uploads last 24 hours and can be claimed into a workspace with krowk claim.
Command
What it does
krowk push <file...>
Upload files and print a link for each
krowk runs / artifacts
Group runs; list, show, attach and delete artifacts (push is artifacts create, short)
krowk login / whoami
Sign in to a workspace (--token for CI)
krowk config set workspace <ws>
Pin a repository to a workspace
Useful push flags include --private, --caption, --link, --metadata key=value, and --destination github|slack|linear|…, which prints the paste format that tool expects. Metadata is public, so never put a secret in it.
It outputs urls, markdown (ready to post as a PR comment), run-slug and json, and adds the links to the job summary.
Using krowk from other agents
skills/krowk/SKILL.md teaches another agent how to use krowk. The installer adds it to ~/.claude/skills.
In Claude Code, /plugin marketplace add krowkcom/krowk and then /plugin install krowk@krowk add the skill and the MCP server together. When krowk drives Claude Code itself, the plugin's server stays off and krowk's own tools are used.
Output is JSON when piped. krowk help --json describes every command, and --jq filters the output without needing jq installed: URL=$(krowk push shot.png --jq '.data.artifacts[0].url').
krowk-mcp serves the same features over MCP stdio, for any other MCP client, or for Claude Code without the plugin: claude mcp add krowk -- krowk-mcp.
Exit codes: 0 ok · 1 bad command · 2 not found · 3 needs credentials · 4 refused · 5 rate limited · 6 transfer failed · 7 server error · 8 gone.
Configuration
Everything krowk keeps is in ~/.krowk/ (move it with KROWK_HOME): settings in config.json, keys in credentials.json, and sessions in sessions/. A repository can commit .krowk/config.json to pin a workspace or add permission rules.
Variable
Purpose
KROWK_TOKEN
API token (takes precedence over stored credentials)
KROWK_WORKSPACE
Which stored workspace key to use
KROWK_API_URL
Point at a self-hosted registry
KROWK_NO_UPDATE_CHECK
1 disables release checks
MCP servers
krowk's own agent uses the MCP servers you already set up for Claude Code: mcpServers in ~/.claude.json (user and per-project) and in Claude's settings.json, plus mcpServers in ~/.krowk/config.json. A repository's .mcp.json is used only after you trust the repository, and your own servers of the same name win over it. Each server takes Claude Code's shape, with ${VAR} and ${VAR:-default} expanded:
A stdio server: the program krowk starts, and its arguments and environment
type: "http", url, headers
A streamable-HTTP server
The model sees two tools, mcp_search and mcp_call, however many MCP tools there are, so they cost no context until one is used. No server starts until the first search. Permission rules name a tool as Mcp(server:tool) or mcp__server__tool, and Mcp(server) names all of a server's tools; a server denied whole is never started.
Development
make check # clippy, unit tests and golden tests
make dev # fast build, linked into ~/.cargo/bin
make mock # local registry; run any command with --dev
Publishing from Krowk, the coding agent harness, for any MCP client: an MCP server that turns an agent's screenshots, recordings, diffs and logs into links that unfurl in Slack and Basecamp; in GitHub
Publishing from Krowk, the coding agent harness, for any MCP client: an MCP server that turns an agent's screenshots, recordings, diffs and logs into links that unfurl in Slack and Basecamp; in GitHub and Linear, images show inline.
ตรวจสอบล่าสุดเมื่อ 2026-11-11 Publishing from Krowk, the coding agent harness, for any MCP client: an MCP server that turns an agent's screenshots, recordings, diffs and logs into links that unfurl in Slack and Basecamp; in GitHub and Linear, images show inline.