JetBrains/thinkrail is a desktop and mobile client that embeds the pi coding agent inside an interface for editing code and managing development workspaces. Developers use it to work on Git repositories through separate worktrees, a Monaco editor, terminals, Git views, specifications, and concurrent agent sessions. The catalogue entries are skills and instructions for its agent-based development workflow.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/JetBrains/thinkrailWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/jetbrains/thinkrail/spec)<a href="https://agentmods.dev/commands/jetbrains/thinkrail/spec"><img src="https://agentmods.dev/badge/commands/jetbrains/thinkrail/spec/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/jetbrains/thinkrail/spec"><img src="https://agentmods.dev/badge/commands/jetbrains/thinkrail/spec.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01337 |
| Opus 5.5 | $0.00000 | $0.00535 |
| Sonnet 5.5 | $0.00000 | $0.00267 |
| Haiku 4.5 | $0.00000 | $0.00134 |
Grade A, and why
SPEC scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Responsibility
The typed contract an invocable app action is described with, so the search palette can list, gate and
run the same actions the menus and chords run. It owns no handlers: a command's run is supplied by
whoever composes the catalog, today shell.
Boundary
- Public surface:
Command,CommandGroup,CommandShortcut,COMMAND_GROUP_LABELS,isCommandEnabled,shortcutKeys. - Allowed deps:
lib(platform family),@thinkrail/locales(the group-labelMessageKeys),@remixicon/reacttypes. - Forbidden: importing
store,transport,panels,chat,shellorsearch; binding keys. A command'sshortcutis display-only — the chord keeps its single owner (useGlobalHotkeys,shortcutCommands, the native menu, or a feature's own key map) and the palette calls the same handler.
Contract
- A command exists only where it applies: the composer builds the catalog for the current context (no
active workspace, no workspace commands), so
runtakes no arguments and re-checks nothing.mutatesmarks a host mutation, whichisCommandEnabledrefuses while offline — the transport replays pending requests on reconnect, so a mutation is never issued blind.disabledReason(already translated) lists a command that applies but cannot run right now, with the reason its owning surface shows; the offline reason wins over it. titleis already in the user's language: the composer builds the catalog withuseT()'s translator (so the palette matches what the user reads), andCOMMAND_GROUP_LABELScarries the group keys it translates the same way.shortcutKeysrenders macOS glyphs in the system order (⌃⌥⇧⌘) andCtrl/Alt/Shiftelsewhere.
Coverage
Every existing action, where it lives, and whether the palette runs it. "Row UI" marks actions that keep their own surface because they need a selection, an inline editor, or a confirmation the palette does not reproduce.
| Command | Handler | Context / gate | Palette |
|---|---|---|---|
workspace.new New workspace… (Mod+N; Mod+Alt+N shown in a browser) |
shell NewWorkspaceDialog |
context project | yes |
workspace.openDefault Work in project folder |
panels/defaultWorkspace.enterDefaultWorkspace |
context project | yes |
chat.new New chat |
panels/openChat.startChatInTab |
active workspace, online | yes |
project.open Open project… |
panels/useOpenProject.pickAndOpen (dialogs mounted by shell) |
online | yes |
project.openPath Open a path on the host… |
panels/useOpenProject.enterHostPath |
online | yes |
project.create Start a new project… |
panels/useOpenProject.startNewProject (CreateProjectDialog; a setup intent starts the setup chat) |
a host serving project.create (v94), online |
yes |
| Reopen a recent project | AddProjectMenu recents |
a recent path | Row UI |
chat.history Search chat history (Ctrl+R) |
shell/chordActions.openChatHistory |
a chat tab | yes |
view.show.{specs,files,review,plan} (every LayoutToolId but projects) |
store.requestToolView |
active workspace | yes |
view.toggleLeft/Right/Bottom (Mod+B/Mod+J/Mod+Shift+J) |
shell/chordActions (the functions the chords run) |
right/bottom need a workspace | yes |
settings.open and every Settings section (incl. Keyboard's double-Shift row, Extensions, MCP servers) |
store.openSettings(section); the native menu's Mod+, is shown where it exists |
protocol-gated sections only | yes |
| Theme, extension and MCP server switches | their Settings sections — the palette opens them, never flips them in place | — | Settings |
workspace.markRead / markUnread |
panels/workspaceActions |
supportsWorkspaceReadState + selectWorkspaceReadAction |
yes |
workspace.settle / unsettle |
panels/workspaceActions |
canSettleWorkspace; Settle disabled while an agent works or a chat needs you (selectWorkspaceSettleBlocked), as in the tree |
yes |
workspace.reveal, workspace.copyPath, workspace.openIn.<editor> |
panels/workspaceActions, lib.copyText |
active workspace | yes |
terminal.new New terminal |
store.addTerminal |
active workspace, online (placement reserves on the host) | yes |
| Remove workspace, close project, stop trusting, layout apply/reset, clear review, MCP approve/remove, stop all subagents, delete template | their confirm dialogs | confirmation | Row UI |
| Publish / push / open, mark ready, merge, reopen a pull request, delete the remote branch, hand checks or a diverged branch to the agent | shell/locationBar topbar review card |
branch status + GitHub facts | Row UI |
| Rename workspace/project/chat | inline editors | an inline editor | Row UI |
| Split/move/fold tabs and groups, focus adjacent group | Workbench internals | a tab or group | Row UI |
/compact, /model, effort, Stop, resources inspector, PR flows, plan review, review fixes |
lazy ChatView, PlanPane, ChangesReviewPane |
a mounted surface | Row UI |
| Open/revert a file or hunk, review comment actions | selection surfaces | a selection | Row UI |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 70 lines · 0 tokens per session scan A 33ede566669a
SPEC is a command published in the GitHub repository JetBrains/thinkrail (7,763 stars, last pushed yesterday), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 1,337 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-02.
SPEC compared with commands from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| SPECJetBrains/thinkrail | Command | 7.8k | yesterday | Apache-2.0 | A | 0 | Claude Code |
| test-featureLykhoyda/rn-dev-agent | Command | 20 | 12d ago | MIT | A | 30 | Claude Code |
| nav-graphLykhoyda/rn-dev-agent | Command | 20 | 12d ago | MIT | A | 21 | Claude Code |
| check-envLykhoyda/rn-dev-agent | Command | 20 | 12d ago | MIT | A | 21 | Claude Code |
| run-workflowLykhoyda/rn-dev-agent | Command | 20 | 12d ago | MIT | A | 49 | Claude Code |
| observeLykhoyda/rn-dev-agent | Command | 20 | 12d ago | MIT | A | 21 | Claude Code |
| rn-feature-devLykhoyda/rn-dev-agent | Command | 20 | 12d ago | MIT | A | 26 | Claude Code |