Organize TeamHarness project and task files in shared workspaces, synchronize changes, and submit deliverables. It also guides publishing files to Matrix chat rooms and explaining file access through other requester channels.
Use for TeamHarness shared workspace paths, filesync, artifact publication, and Matrix room file boundaries, including artifact publish_file and task/project deliverables. Do not use for message routing or project state changes.
Review the instructions the agent will receive before you install.
Use shared workspace paths in task specs and team messages.
Project files belong under shared/projects/{project-id}/. Task specs,
deliverables, and results belong under shared/tasks/{task-id}/.
Do not expose object storage internals in human-facing messages. Use TeamHarness filesync tools for explicit shared file operations.
shared/... paths are durable collaboration state. They let Leaders, Workers,
and recovery flows find the same project/task files.
Matrix room files are UI-visible file events. The room Files panel is populated
from Matrix m.file events, not from text that merely mentions a shared/...
path.
DingTalk, Feishu, WeChat, and other non-Matrix requester channels do not receive
TeamHarness artifact publish_file room files. For those requester reports,
mention the shared/... artifact paths and say they are available from the
shared workspace or platform object-storage view. Do not tell the requester to
open a chat file card unless the file was actually published to a Matrix room.
For important task outputs, write the files under shared/tasks/{task-id}/ and
list them in taskflow submit_task deliverables. Worker task deliverables
must not include shared/projects/... paths. For important project reports,
the Leader writes shared/projects/{project-id}/result.md from accepted
project state before publishing it.
Use artifact publish_file when an explicit extra workspace file must become a
Matrix room file outside the task/project submit flow. The path can be any
workspace-relative file path, not only shared/..., but it must not be absolute
or escape the workspace. Always pass an explicit Matrix room target or
roomId. When publishing files for a requester report that was just sent with
message to a Matrix room, pass the returned Matrix messageId as
parentEventId.
Do not manually send sensitive files. TeamHarness refuses obvious sensitive
paths and text content by default, and reports the publish status in
publishedArtifacts.
Use these paths in task specs and team messages:
shared/projects/{project-id}/meta.json
shared/projects/{project-id}/plan.md
shared/projects/{project-id}/result.md
shared/tasks/{task-id}/meta.json
shared/tasks/{task-id}/spec.md
shared/tasks/{task-id}/workspace/
shared/tasks/{task-id}/result.md
The Leader owns project files and task specs. Workers own task workspaces, task deliverables, and task results.
Use filesync when you need an explicit shared file operation.
List a concrete shared directory:
{
"action": "list",
"path": "shared/projects/demo-project-001"
}
Pull before reading remote shared state:
{
"action": "pull",
"path": "shared/tasks/task-001"
}
Push after writing project-level files:
{
"action": "push",
"path": "shared/projects/demo-project-001"
}
Do not ask humans or Workers to inspect storage bucket names, access keys, or
provider-specific prefixes. Use shared/... paths in all visible coordination.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Find worker agents in Nacos that match a task, recommend candidates, and import the administrator's chosen worker. It also supports importing a specified nacos package URI, with worker name confirmation and direct reporting of installation failures.
Prepare an existing OpenClaw setup for import as an AgentTeams managed Worker. Analyze tool dependencies, adapt role instructions and scheduled jobs, and build a ZIP package containing the configuration and custom image requirements.
Manage communication channels by identifying senders, maintaining trusted contacts, and setting the administrator's primary notification destination. It also covers first contact on a new channel and contacting the administrator across channels, with rules for handling unknown senders.
Prepare coding tasks for delegation to an authorized agent’s AI command-line tool. Set up and synchronize the workspace, write precise change requests, then review returned changes and run relevant checks before reporting completion.
Run AI coding command-line tools on behalf of worker agents using their prepared prompts and workspaces. Detect available tools, configure delegation, return results, and apply supported execution controls such as time budgets and permission settings.
Guides communication in Matrix agent teams by choosing recipients, rooms, and when to mention someone. It covers task notifications and progress reports, helping agents avoid duplicate assignments and send concrete updates to the appropriate requester.