agentleFS
es6kres6kr/skills/skills/orca/SKILL.md

orca

Skill2 starsChanged 37 days agoMITView on GitHub

Launch and coordinate agent sessions inside the Orca IDE runtime (stablyai/orca). Topics: send - find a running Orca terminal by title/preview/worktree and deliver a prompt to it, always disambiguating with a question when more than one candidate matches and never targeting the caller's own pane [send.md]. launch - start a new agent session (claude, antigravity, openclaw, codex, and 30+ other supported agents) in a fresh or existing worktree [launch.md]. install - install the stablyai/orca skill bundle into Claude Code, OpenClaw, or Antigravity, asking whether to ghq-clone + symlink or use a plain marketplace install [install.md]. Use when: "send this to the claude session working on X", "hand this off to another agent in Orca", "spawn a new codex/antigravity worktree", "which Orca terminal is running Y", "install the orca skills into openclaw/antigravity", "Orca worktree", "Orca terminal send".

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every topic in this skill needs the Orca CLI executable.

    line 40
  2. Very specific:

    If ORCA_CLI_COMMAND is set, use its value (Orca exports this for managed WSL sessions).

    line 43
  3. Distinctive:

    If the resolved executable fails, report the exact error and stop.

    line 53
  4. Very specific:

    If result.app.running is false, start it with ORCA open --json first.

    line 62
  5. Very specific:

    Every command in this skill that talks to another terminal must first know which terminal it is itself, to avoid ever targeting its own …

    line 84

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Orca Session Management

Coordinates agent sessions running inside the Orca IDE (stablyai/orca) — finding a specific running session and delivering work to it, launching new sessions, and installing Orca's own bundled skills into other harnesses.

Topic Dispatch

When this skill is invoked with a topic specifier (e.g., /orca send or Skill("orca", "send")), load and follow only the matching topic file (send.md). Do not echo the Topics table or summarize other topics in the response. The Topics table below is an index for invocations without a topic specifier — it is not user-facing output when a topic is named.

Topics

TopicDescriptionGuide
sendDeliver work to a running Orca terminal, disambiguating candidatessend.md
launchStart a new agent session in a worktreelaunch.md
installInstall the Orca skill bundle into another harnessinstall.md

Resolve the CLI (all topics)

Every topic in this skill needs the Orca CLI executable. Resolve it once per session and reuse it:

  1. If ORCA_CLI_COMMAND is set, use its value (Orca exports this for managed WSL sessions).
  2. Otherwise, in a dev checkout exposing ORCA_DEV_REPO_ROOT, use orca-dev.
  3. Otherwise, on Linux outside an Orca-managed terminal, use orca-ide. Never run bare orca there — it normally resolves to the GNOME Orca screen reader and starts speech on the user's machine.
  4. Otherwise, use orca.

Below, ORCA is a placeholder for the resolved executable. Substitute it before running anything — do not create a shell variable or run ORCA literally.

If the resolved executable fails, report the exact error and stop. Do not fall through to another executable — that could silently target a different Orca build.

Confirm the app is reachable before anything else:

ORCA status --json

If result.app.running is false, start it with ORCA open --json first. Prefer --json for every agent-driven call.

Command source of truth

Do not hardcode Orca CLI flags in this skill or memorize them from a prior session. Orca releases change subcommands and flags; a cached copy drifts from the binary that will actually run the command. Before running an unfamiliar command, confirm its current shape with:

ORCA skills get orca-cli
ORCA skills get orchestration

These print the complete, version-matched guide for the exact binary in this session. This skill's topic files hold the parts that don't change across releases (the disambiguation policy, the self-targeting guard, the harness install matrix) — not a copy of Orca's flag reference.

Self-identification

Every command in this skill that talks to another terminal must first know which terminal it is itself, to avoid ever targeting its own pane. Orca exports ORCA_PANE_KEY as <tabId>:<leafId> inside every managed terminal. If it is unset, this session is not running inside an Orca-managed terminal and send must refuse to run (see send.md).

Core Terminal Lifecycle & Layout Operations

While send.md and launch.md detail higher-level handoff workflows, Orca CLI provides direct pane and tab manipulation commands:

CommandPurposeExample
splitSplit active pane into paired panesORCA terminal split --terminal <handle> --direction vertical --command "<cmd>" --json
renameUpdate the UI tab titleORCA terminal rename --terminal <handle> --title "<title>" --json
switchFocus/switch UI view to terminalORCA terminal switch --terminal <handle> --json
closeGracefully terminate a terminal/tabORCA terminal close --terminal <handle> --json
waitAwait condition (e.g. idle prompt)ORCA terminal wait --terminal <handle> --for tui-idle --timeout-ms 60000 --json

Topic Dependencies

orca
  ├─→ send: uses the CLI-resolution + self-identification rules above
  ├─→ launch: uses the CLI-resolution rules above
  └─→ install: independent — only needs the CLI resolved when verifying a Claude/OpenClaw install
  • send and launch both call into the running Orca app; install mostly edits files on disk and only touches the CLI to confirm an install landed.
  • For task DAGs, ask/reply flows, or coordinator loops between agents, use Orca's own orchestration skill directly (ORCA skills get orchestration) — this skill does not reimplement it. send covers the simpler case: "deliver this prompt to that terminal."

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.