Paseo
Paseo reference for managing hosts, projects, workspaces, workspace scripts, agents, schedules, and heartbeats.
mkdir -p ~/.claude/skills/paseo && curl -fsSL 'https://raw.githubusercontent.com/getpaseo/paseo/main/skills/paseo/SKILL.md' -o ~/.claude/skills/paseo/SKILL.md
Get the complete skill folder, including its scripts and reference files.
When to use this skill
Use Paseo when an AI agent needs a reusable SKILL.md workflow for this job: Paseo reference for managing hosts, projects, workspaces, workspace scripts, agents, schedules, and heartbeats.
When not to use it
Skip Paseo when the task is outside the productivity category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.
How to install
Where will you use it?
mkdir -p ~/.claude/skills/paseo && curl -fsSL 'https://raw.githubusercontent.com/getpaseo/paseo/main/skills/paseo/SKILL.md' -o ~/.claude/skills/paseo/SKILL.md
- Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/paseo/SKILL.md. For a shared project, commit the folder at .claude/skills/paseo/ instead.
- Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
- For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.
Full install guide for Claude, Cursor, and Codex →
Try a starter prompt
After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.
Use the Paseo skill for this task: Use Paseo when this documented workflow matches the task. Confirm any required inputs or files before making changes.
What supports this recommendation?
68/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.
- Documentation
- 71/100
- Practical detail
- 68/100
- Source evidence
- 63/100
- Source signals
- 70/100
Model compatibility evidence
- claudeinferred
The skill text mentions Claude or a closely associated term. This is keyword-derived compatibility; no hands-on model test is recorded.
- chatgptinferred
The skill text mentions ChatGPT or a closely associated term. This is keyword-derived compatibility; no hands-on model test is recorded.
- geminiuntested
No model-specific signal or recorded compatibility test was found.
- copilotuntested
No model-specific signal or recorded compatibility test was found.
- llamauntested
No model-specific signal or recorded compatibility test was found.
- perplexityuntested
No model-specific signal or recorded compatibility test was found.
- mistraluntested
No model-specific signal or recorded compatibility test was found.
- grokuntested
No model-specific signal or recorded compatibility test was found.
Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.
What this skill does
Paseo is a remote daemon that manages coding agents, terminals. Control it through MCP tools or the CLI.
## Connected hosts
Call `list_hosts` to discover the calling daemon (`self`) and its registered destinations (`hosts`). Use a destination's `serverId` for tool calls; labels are display names.
`create_workspace`, `list_workspaces`, `create_agent`, `send_agent_prompt`, `get_agent_status`, `get_agent_activity`, and `list_agents` accept `serverId`. Omit it for the calling daemon. Other MCP tools operate on the calling daemon.
Keep an agent's `{ serverId, agentId }` together. For example, call `send_agent_prompt` with:
```json { "serverId": "srv_build", "agentId": "agent_123", "prompt": "Continue", "notifyOnFinish": true } ```
Remote paths and workspace IDs belong to the destination. Remote `list_agents` has no caller-directory default; pass `cwd` to filter it.
Manage registrations through the CLI; use the returned `registrationId` for updates and removal:
```bash paseo --host localhost:6768 host add localhost:6779 --label build paseo --host localhost:6768 host ls --json ```
## Projects
Manage the daemon's project registry through the CLI:
```bash paseo project create [path] paseo project ls paseo project rename <project-id> <name> paseo project rename <project-id> --reset paseo project delete <project-id> ```
For a local daemon, `project create` defaults to the current directory and resolves relative paths on the CLI machine. With `--host` or `PASEO_HOST`, always provide a path; the target daemon interprets it on its own machine. Deleting a project archives its active workspaces and removes the project from Paseo without deleting the project directory.
## Workspaces
**`create_workspace`** — create a workspace independently of any agent. Required: `isolation` (`local` or `worktree`). Worktree isolation supports `mode: "branch-off" | "checkout-branch" | "checkout-pr"`: use `branchName`/`baseBranch` for a new branch, `branch` for an existing branch, or `prNumber` plus optional `forge` for a change request. Use `path` for a local directory or source checkout and `projectId` for the owning project; when no path is supplied, a worktree may resolve its source from `projectId`. `worktreeSlug` controls the managed path. Returns the workspace descriptor centered on `workspaceId`.
`background: true` hides a workspace from default lists and the sidebar while its contents remain durable and accessible; omit `background` to inherit your current workspace’s value (false without a caller), and override it only when the user explicitly asks.
Choose `baseBranch` explicitly: `origin/main` selects the remote-tracking branch; `refs/heads/main` selects local main. Bare `main` prefers local main when it exists, otherwise origin/main. Paseo retains the resolved ref for workspace comparisons, even after rebasing the branch or changing its PR target.
**`list_workspaces`** — list active workspaces, excluding background workspaces unless `includeBackground: true`.
**`archive_workspace`** — `{ workspaceId }`. Archives the workspace, its agents, and its terminals. Local directories remain; Paseo removes an owned worktree only after its final active workspace reference is archived.
**`rename_workspace`** — `{ workspaceId, name }`. Rename workspace.
## Workspace scripts
Configured `paseo.json` scripts use the same supervised lifecycle from tools and the CLI.
**`list_workspace_scripts`** — `{ workspaceId }`. Lists configured scripts with lifecycle, service port, proxy URLs, health, exit code, and terminal ID.
**`start_workspace_script`** — `{ workspaceId, scriptName }`. Starts one configured script through Paseo's managed workspace-script launcher and returns its status metadata.
**`stop_workspace_script`** — `{ workspaceId, scriptName }`. Stops a running script through its supervised terminal and returns the stopped status metadata.
The matching CLI surface accepts either an explicit workspace ID or resolves the current directory:
```bash paseo script ls [--cwd <path> | --workspace <workspace-id>] paseo script start <name> [--cwd <path> | --workspace <workspace-id>] paseo script stop <name> [--cwd <path> | --workspace <workspace-id>] ```
## Agents
**`create_agent`** — required: `title`, `provider` (`claude/opus`, `codex/gpt-5.4`, …), `initialPrompt`. Optional: `workspaceId`, `notifyOnFinish`, `settings`, `labels`. Returns `{ agentId, workspaceId, … }`.
Initial runtime settings live under `settings`: `modeId`, `thinkingOptionId`, and provider-specific `features`. Agent profiles are the preferred source for these values. For Codex fast mode, pass `settings: { features: { "fast_mode": true } }` when creating the agent.
Agent-scoped creation always creates your subagent. On your own host, omit `workspaceId` to use your current workspace. On another host, first list or create a workspace with that `serverId`, then pass its `workspaceId` and the same `serverId` to `create_agent`. Placement never changes parentage.
Detach is an explicit user action in the subagents track, not an agent tool. A cross-workspace child remains your subagent even though it also appears as a normal tab in its workspace.
Agent-scoped `create_agent` defaults `notifyOnFinish` to true. Set it to `false` when you want explicit reporting or no completion callback.
**`send_agent_prompt`** — `{ agentId, prompt, activeTurnBehavior? }`. Use for follow-ups to an existing agent. Agent-scoped calls default to `steer`; top-level calls default to `interrupt`. For agent-scoped calls, `steer` waits until the recipient's current turn ends, and only `interrupt` stops a running turn. For top-level calls, `steer` joins the running turn and falls back to interruption when unavailable. CLI agent sends use the same defaults and accept `--behavior steer|interrupt`. Agent-scoped calls return a queued receipt immediately; messages to each recipient deliver together after two quiet seconds, or after the recipient's turn ends if it is running. Top-level calls return after dispatch acknowledgement. Agent-scoped calls default `notifyOnFinish` to true; top-level calls have no parent callback.
**`update_agent`** — `{ agentId, name?, labels?, settings? }`. Use `settings` for runtime changes on an existing agent: `modeId`, `model`, `thinkingOptionId`, and provider-specific `features`. For Codex fast mode, pass `settings: { features: { "fast_mode": true } }`.
**`list_agents`** — filter by `cwd`, `statuses`, `sinceHours`, `includeArchived`; set `includeBackground: true` to include agents in background workspaces.
**`archive_agent`** — `{ agentId }`. Interrupts if running, removes from active list.
## Agent profiles and provider discovery
**`list_profiles`** — named launch bundles configured by the human. Before choosing how to launch a delegated agent, call this tool and read every profile's `notes`. Pick a named profile the user requested, or the profile whose notes best match the work.
There is no `profile` parameter on `create_agent`. Materialize the selected profile into the call:
- combine `provider` and `model` as the `provider/model` value for `create_agent.provider` - copy `modeId` to `settings.modeId` - copy `thinkingOptionId` to `settings.thinkingOptionId` - copy `featureValues` to `settings.features`
Omit absent values. Do not remember a selected profile or infer drift later; a profile is only launch configuration.
If no profile fits, or no profiles are configured, use the provider discovery tools below rather than guessing. Tell the user when you fall back because no configured profile fits.
**`list_providers`** — compact provider availability and modes.
**`list_models`** — full model list for one provider. Use only when you need model IDs or thinking options; the list can be large.
**`inspect_provider`** — compact provider capability and feature inspection. Required: `provider`; pass `cwd` when you are not in an agent-scoped session. Optional: `settings` with draft `model`, `modeId`, `thinkingOptionId`, and `features`.
Only set feature IDs returned by `inspect_provider`. For Codex fast mode, look for `fast_mode` and pass `settings: { features: { "fast_mode": true } }` to `create_agent` or `update_agent`.
## Schedules and heartbeats
**`create_schedule`** — starts a new agent on a cron cadence. Required: `prompt`, `cron`, `provider`. Optional: `timezone`, `name`, `cwd`, `maxRuns`, `expiresIn`. Use when the recurring work should live in fresh agents.
**`create_heartbeat`** — sends you a prompt on a cron cadence. Required: `prompt`, `cron`. Optional: `timezone`, `name`, `maxRuns`, `expiresIn`. Use for reminders, PR/build babysitting, and status checks that should return to this conversation.
**`delete_heartbeat`** stops it. MCP intentionally exposes no heartbeat update tool; delete and recreate when its task or cadence changes.
Schedules have the full list/inspect/update/pause/resume/run-once/log/delete surface. Heartbeats deliberately do not.
## Waiting
When you create or prompt a subagent, `notifyOnFinish` defaults to `true` and reports completion, errors, or permission requests, including across hosts. Finish your turn and wait for that notification instead of polling. If you want an explicit reply with `send_agent_prompt`, have the sender check `list_hosts` for the recipient’s server before setting `notifyOnFinish: false`.
## CLI semantics
The CLI and tools use the same ownership semantics even where their syntax differs:
```bash paseo workspace create --isolation worktree --mode branch-off --new-branch fix-x --base origin/main paseo workspace create --isolation worktree --mode checkout-branch --branch existing-work paseo workspace create --isolation worktree --mode checkout-pr --pr-number 42 paseo run --provider codex/gpt-5.4 --mode full-access --workspace <workspace-id> "<prompt>" paseo run --provider codex/gpt-5.4 --mode full-access --new-workspace worktree --worktree-mode branch-off --new-branch fix-x --base origin/main "<prompt>" paseo send <agent-id> "<follow-up>" paseo ls paseo schedule create --cron "*/15 * * * *" "ping main build" paseo heartbeat create --cron "*/15 * * * *" "check the build" ```
Discover with `paseo --help` and `paseo <cmd> --help`.
For product questions, setup, logs, version problems, or troubleshooting, use the **paseo-help** skill.
Intended uses
- Use Paseo when this documented workflow matches the task.
Implementation notes
- Review the source instructions and adapt inputs before running the workflow.
Try it, then tell us how it went
A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.