EVC Mesh Task Management

OpenClaw skill for EVC Mesh — a task management platform for coordinating humans and AI agents. Source: evc-mesh-openclaw-skill.

Environment Variables

Variable	Required	Description
MESH_API_URL	Yes	Base URL of the Mesh API (e.g. https://your-mesh-instance.example.com)
MESH_AGENT_KEY	Yes	Agent API key (agk_workspace_...) from Mesh admin panel

All scripts read these from the environment. No separate auth step needed —
the agent key is passed as X-Agent-Key header on every request.

Quick Start

All scripts live in the scripts/ directory of this skill. When running from
an OpenClaw agent, use the full path to avoid "command not found" errors:

# Set the skill directory (adjust if installed elsewhere)
SKILL_DIR="/opt/openclaw/skills/evc-mesh"

export MESH_API_URL="https://your-mesh-instance.example.com"
export MESH_AGENT_KEY="agk_your_workspace_key"

# Send heartbeat (register as online)
bash "$SKILL_DIR/scripts/heartbeat.sh" online

# Auto-discover workspace, projects, statuses
bash "$SKILL_DIR/scripts/discover.sh"

# Or list projects (auto-resolves workspace from agent key)
bash "$SKILL_DIR/scripts/list-projects.sh"

# List statuses for a project (to know status IDs)
bash "$SKILL_DIR/scripts/list-statuses.sh" <project_id>

# List tasks assigned to you
bash "$SKILL_DIR/scripts/list-tasks.sh" <project_id> --assignee me

# Get task details
bash "$SKILL_DIR/scripts/get-task.sh" <task_id>

# Move task to in_progress
bash "$SKILL_DIR/scripts/move-task.sh" <task_id> <in_progress_status_id>

# Add a comment
bash "$SKILL_DIR/scripts/add-comment.sh" <task_id> "Starting work on this task"

# Upload an artifact
bash "$SKILL_DIR/scripts/upload-artifact.sh" <task_id> "report" "log" ./output.log

# Delete a test task
bash "$SKILL_DIR/scripts/delete-task.sh" <task_id>

# Move task to done when finished
bash "$SKILL_DIR/scripts/move-task.sh" <task_id> <done_status_id>

Tip: You can also cd "$SKILL_DIR" first and then use relative paths
like bash scripts/heartbeat.sh online.

Session Start Protocol

At the beginning of every work session:

1. Discover environment: bash scripts/discover.sh --export — gets agent ID, workspace, projects, statuses in one call
2. Send heartbeat: bash scripts/heartbeat.sh online
3. (Optional) List agents: bash scripts/list-agents.sh — see other agents in workspace for coordination
4. Check assigned tasks: bash scripts/my-tasks.sh — shows all tasks assigned to you
5. Read context: Check recent comments and activity on your tasks
6. Check triage inbox: bash scripts/list-triage.sh <workspace_id> — pick up unrouted tasks

If there are tasks in in_progress from a previous session, resume those first.
If all assigned tasks are todo, pick the highest-priority one.

Priority order: urgent > high > medium > low.

Task Pickup Protocol

1. Get full context: bash scripts/get-context.sh <task_id> — returns task + status + comments + deps + artifacts in one call
2. Check dependencies: if any blocker is not done, skip this task
3. Assign to yourself: bash scripts/assign-task.sh <task_id> <your_agent_id>
4. Move to in_progress: bash scripts/move-task.sh <task_id> <in_progress_status_id>
5. Comment your plan: bash scripts/add-comment.sh <task_id> "Starting: <brief plan>"
6. Send heartbeat: bash scripts/heartbeat.sh busy

During Work

• Post progress comments at meaningful checkpoints (not every action)
• If the task is larger than expected, create subtasks: bash scripts/create-subtask.sh <task_id> "Subtask title"
• Upload artifacts when files are ready: bash "$SKILL_DIR/scripts/upload-artifact.sh" <task_id> <name> <type> <file_or_content>
• Link PRs, commits, or branches: bash "$SKILL_DIR/scripts/link-vcs.sh" <task_id> pr <pr_number> <pr_url> --title "Fix auth bug"
• Check custom fields if needed: bash "$SKILL_DIR/scripts/list-custom-fields.sh" <project_id>

Task Completion Protocol

1. Upload final artifacts
2. Link VCS references: bash scripts/link-vcs.sh <task_id> pr <number> <url>
3. Add completion comment: bash scripts/add-comment.sh <task_id> "Done. <summary of what was done>"
4. Move to done (or review): bash scripts/move-task.sh <task_id> <done_status_id>
5. Send heartbeat: bash scripts/heartbeat.sh online

Project Status Updates

After completing a milestone or sprint, post a project update:

bash scripts/post-update.sh <project_id> "Sprint 5 Complete" "Shipped auth module and fixed 12 bugs" --status on_track

Error Handling

If you encounter an unrecoverable error:

1. Comment the error: bash scripts/add-comment.sh <task_id> "Blocked: <error description>"
2. Move back to todo: bash scripts/move-task.sh <task_id> <todo_status_id>
3. Unassign: bash scripts/assign-task.sh <task_id> --unassign
4. Send heartbeat: bash scripts/heartbeat.sh error

Task Creation

When you identify untracked work:

bash scripts/create-task.sh <project_id> "Add pagination to /tasks endpoint" \
  --priority high \
  --description "The endpoint returns all tasks without limit"

Recurring Tasks

Set up automated recurring tasks that create new instances on a schedule. Each instance gets context from the previous run.

# Create a weekly code review schedule assigned to an agent
bash scripts/create-recurring-schedule.sh --project-id <id> \
  --title-template "Weekly Code Review — {{.Date}} (#{{.Number}})" \
  --frequency weekly --assignee-id <agent_id> --assignee-type agent \
  --priority high --timezone "Europe/Moscow"

# List active recurring schedules
bash scripts/list-recurring-schedules.sh --project-id <id>

# View history of a recurring schedule (previous instances + comments)
bash scripts/get-recurring-history.sh --schedule-id <id> --limit 5

# Trigger next instance immediately (don't wait for schedule)
bash scripts/trigger-recurring-now.sh --schedule-id <id>

When you receive a recurring task, always check the history first to understand what previous instances accomplished.

Multi-Agent Coordination

Agents coordinate through shared task state — not direct communication.

• Always check assigned tasks before picking up new ones
• Never work on a task assigned to another agent
• Post comments and summaries so other agents can read context
• Check recent comments/activity before modifying shared files

Script Reference

Script	Purpose	Args
whoami.sh	Get agent profile (id, workspace_id)	(no args)
heartbeat.sh	Register agent status	<status> (online/busy/error)
my-tasks.sh	List tasks assigned to me	(no args)
discover.sh	Auto-discover workspace, projects, statuses	[--export] [--json]
list-projects.sh	List workspace projects	[workspace_id] (auto-resolves if omitted)
list-agents.sh	List agents in workspace	[workspace_id] (auto-resolves if omitted)
get-project.sh	Get project details	<project_id>
list-statuses.sh	List project statuses	<project_id>
list-custom-fields.sh	List custom field definitions	<project_id>
list-tasks.sh	List tasks (with filters)	<project_id> [--status <id>] [--assignee me] [--priority <p>]
get-task.sh	Get task details	<task_id>
get-context.sh	Get enriched task context	<task_id>
create-task.sh	Create a task	<project_id> <title> [--priority p] [--description d] [--status id] [--due_date d] [--estimated_hours n] [--labels l1,l2] [--assignee id] [--parent_task_id id]
update-task.sh	Update task fields	<task_id> [--title t] [--priority p] [--description d] [--due_date d] [--estimated_hours n] [--labels l1,l2] [--assignee id]
move-task.sh	Change task status	<task_id> <status_id>
assign-task.sh	Assign/unassign task	<task_id> <agent_id> | --unassign
add-comment.sh	Add comment to task	<task_id> <content>
list-comments.sh	List task comments	<task_id>
create-subtask.sh	Create subtask	<parent_task_id> <title> [--priority p] [--description d] [--due_date d] [--estimated_hours n] [--labels l1,l2] [--assignee id] [--status id]
list-dependencies.sh	List task dependencies	<task_id>
link-vcs.sh	Link PR/commit/branch	<task_id> <type> <external_id> <url> [--title t] [--provider p]
list-vcs-links.sh	List VCS links on task	<task_id>
upload-artifact.sh	Upload artifact	<task_id> <name> <type> <file_or_content>
list-artifacts.sh	List task artifacts	<task_id>
delete-task.sh	Delete a task	<task_id>
publish-event.sh	Publish event to bus	<project_id> <event_type> <subject> <payload_json>
list-events.sh	List project events	<project_id>
get-activity.sh	Get task activity log	<task_id>
post-update.sh	Post project status update	<project_id> <title> <summary> [--status s]
list-triage.sh	List triage inbox tasks	<workspace_id>
get-team.sh	Get workspace team directory (agents + humans)	[workspace_id] (auto-resolves if omitted)
get-workflow-rules.sh	Get workflow rules + caller permissions for project	<project_id>
get-assignment-rules.sh	Get effective assignment rules for project	<project_id>
update-agent-profile.sh	Update calling agent's profile fields	[--role r] [--capabilities go,react] [--zone Backend] [--escalation-to id] [--accepts-from id1,id2] [--max-tasks n] [--hours "24/7"] [--description d]
import-config.sh	Import workspace config from YAML file	<path-to-yaml-file> [workspace_id]
export-config.sh	Export workspace config as YAML	[workspace_id] [--output file.yaml]
create-recurring-schedule.sh	Create a recurring task schedule	--project-id <id> --title-template <t> --frequency daily|weekly|monthly|custom [--cron-expr e] [--timezone tz] [--assignee-id id] [--assignee-type agent|user] [--priority p] [--labels l1,l2]
list-recurring-schedules.sh	List recurring schedules for project	--project-id <id> [--active-only] [--no-active-only]
get-recurring-history.sh	Get instance history for schedule	--schedule-id <id> [--limit n] [--page n]
trigger-recurring-now.sh	Trigger recurring schedule immediately	--schedule-id <id>