canned-response

A canned response is a catalog resource that stores a reusable pair: a message body and a persona override. Pick one in a running agent’s follow-up box to fill in the follow-up text and set the persona for that message in a single click. Use it for replies you send repeatedly, such as “rebase and re-run the tests” or “looks good, open the PR.” A canned response is either tenant-level or workspace-scoped, decided by its name. A bare name (rebase-retest) is available in the workspaces its workspace_use policy admits; a prefixed name (backend/rebase-retest) is usable only in that workspace. Read permissions independently control visibility. Manage them under Organization Settings → Canned Responses.

Fields

Name	Type	Required	Description
`workspace_use`	object	no	For tenant-level names, which workspaces may use this resource: `scope` is `WORKSPACE_USE_SCOPE_NONE` (default), `WORKSPACE_USE_SCOPE_ALL`, or `WORKSPACE_USE_SCOPE_SELECTED` with a `workspaces` list. Scoped names carry no policy and are usable only in their own workspace. See resource admission.
`name`	string	yes	Either a bare DNS label (`rebase-retest`, tenant-level) or `{workspace}/{slug}` (`backend/rebase-retest`, workspace-scoped). Each segment matches `[a-z][a-z0-9-]{0,62}`.
`text`	string	yes	The message body, inserted verbatim into the follow-up field when the response is picked. Maximum 16 KiB (16384 bytes).
`persona`	string	yes	Persona override applied to the follow-up. A bare name (`murmur-programmer`, resolved against the canned response’s own workspace scope: tenant-level when the name is bare) or a pinned `{workspace}/{slug}` (`backend/programmer`). Must resolve to an existing agent-persona with a non-empty prompt on every write.
`title`	string	no	Human-readable label shown in the picker (e.g. `Rebase and re-run tests`). Display only: never inserted into the message. Maximum 1024 bytes.

Both text and persona are required. A canned response always carries a persona override, so picking one both fills the message and switches the persona for that follow-up.

Examples

Create a tenant-level canned response

name: rebase-retest
title: "Rebase and re-run tests"
text: |
  Please rebase onto the latest main, resolve any conflicts, and
  re-run the full test suite before pushing.
persona: murmur-programmer

cat <<'EOF' | murmur set canned-response rebase-retest
name: rebase-retest
title: "Rebase and re-run tests"
text: |
  Please rebase onto the latest main, resolve any conflicts, and
  re-run the full test suite before pushing.
persona: murmur-programmer
EOF

Create a workspace-scoped canned response

A name with a workspace prefix appears only for agents in that workspace. The prefix workspace must exist.

cat <<'EOF' | murmur set canned-response backend/lgtm
name: backend/lgtm
title: "Approve and merge"
text: "Looks good. Approve the PR and merge once checks are green."
persona: backend/reviewer
EOF

Listing canned responses

murmur get canned-response

backend/lgtm
rebase-retest

The list prints one name per line, with no header.

Reading a single canned response

murmur get canned-response rebase-retest

Errors

Code	Meaning	What to do
`INVALID_ARGUMENT`	`name is required`	Provide a `name`.
`INVALID_ARGUMENT`	`canned response name must be a DNS label or {workspace}/{name}`	Use a bare DNS label or a `{workspace}/{slug}` name; each segment is `[a-z][a-z0-9-]{0,62}`.
`INVALID_ARGUMENT`	`text is required`	Provide the message body in `text`.
`INVALID_ARGUMENT`	`text exceeds 16384 byte limit (<n> bytes)`	Shorten `text` to 16 KiB or fewer.
`INVALID_ARGUMENT`	`title exceeds 1024 byte limit (<n> bytes)`	Shorten `title` to 1024 bytes or fewer.
`INVALID_ARGUMENT`	`persona must be a DNS label or {workspace}/{name}`	Use a bare persona name or a pinned `{workspace}/{slug}`.
`INVALID_ARGUMENT`	`persona "<name>" does not resolve to a usable agent persona`	Reference an existing agent-persona. Create it first if needed.
`INVALID_ARGUMENT`	`workspace "<name>" does not exist`	The workspace prefix names a workspace that does not exist. Create it or fix the name.

agent-persona: the persona a canned response sets on the follow-up
Agent personas: concept overview
Dashboard: where you pick a canned response in the follow-up composer
Workspaces: the scope a prefixed canned response belongs to
murmur set: CLI command for creating and updating catalog resources
murmur get: CLI command for reading catalog resources

Getting Started

Core Concepts

Guides

canned-response

Fields

Examples

Create a tenant-level canned response

Create a workspace-scoped canned response

Listing canned responses

Reading a single canned response

Errors

​Fields

​Examples

​Create a tenant-level canned response

​Create a workspace-scoped canned response

​Listing canned responses

​Reading a single canned response

​Errors

​Related

Fields

Examples

Create a tenant-level canned response

Create a workspace-scoped canned response

Listing canned responses

Reading a single canned response

Errors

Related