aweb-tasks
Skill129 starsChanged 4 days agoMITView on GitHub
Use only when the team has aweb tasks or its instructions say it uses them, for sharing tasks, claiming work, and the ready/active loop. Otherwise use the team's own tracker. Creating the first task is the opt-in.
In this repository10 other files it gives its agents.
- AGENTS.mdAGENTS.md
- code-review-disciplineSkill
- cross-repo-changeSkill
- deploy-awid-siteSkill
- releaseSkill
- aweb-bootstrapSkill
- aweb-coordinationSkill
- aweb-identitySkill
- aweb-messagingSkill
- aweb-team-membershipSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 13
Read aw instructions show and the team's existing work context.
- Very specific: line 79
Before claiming, run aw work active to make sure nobody is already on the same scope.
- Very specific: line 81
When status changes, update the task.
- Very specific: line 87
Update task status and close finished tasks with aw task close <id> --reason "what landed and where validated".
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Use this skill only when the team has aweb tasks or its instructions say it uses them. Otherwise use the team's own tracker. Creating the first task is the opt-in; do that when the user or team deliberately chooses aweb tasks, never merely because aweb messaging or these skills are installed.
aweb Tasks
Choosing the team's tracker
Read aw instructions show and the team's existing work context. If the tracker
is unclear, aw task list is a read-only way to check whether the team has tasks.
Empty aw work ready and aw work active results do not establish that the team
has no tasks: it may have only closed tasks. An unavailable task service is an
error, not evidence of an empty team. Do not create a task just to discover
whether this skill applies.
Ready and active loop
First complete the session loop in aweb-coordination: team status, waiting
mail, then pending chat. If that skill is unavailable, run aw workspace status,
aw mail inbox, and aw chat pending in that order before discovering work.
Then inspect the current team's work:
aw work ready # smallest actionable unclaimed item
aw work active # check who already owns related scope before claiming
aw work blocked # dependencies or external answers
aw task show <task-id> # read comments and dependencies before actingTasks record work and ownership where teammates can see it. Follow the team's assignment policy and keep status and comments current instead of maintaining a private duplicate tracker. This choice is per team; another team may use a different tracker.
Sharing tasks
A task is the durable record of a unit of work. Anyone in the team can see it; it doesn't depend on local notes.
aw task create --title "<title>" --description "<details>"
aw task list # filter with --status, --assignee, --labels
aw task show <task-id>
aw task update <task-id> --status in_progress
aw task comment add <task-id> "validation results, blockers, decisions"
aw task close <task-id> --reason "what landed and where validated"
aw task dep add <task-id> <depends-on-id> # express dependenciesWhen creating a task: keep the scope small enough that one agent can complete it. Put what's known into the body so a teammate can pick it up without asking. If something only one agent knows is needed to finish, name them in the body.
Requesting a human decision
A task assigned to a human is a request. Humans are valid task assignees in the same team even though the agent roster does not list them. Assign the request to the human whose message you are acting on, using that message's alias. Do not guess from a display name or try to discover humans through the roster. State the question, relevant options and what it unblocks, and keep the task open while waiting. Do not mail the human first: agents cannot start a conversation with a human. After creating the request, end your turn: do not poll or sleep waiting for the answer.
The answer usually arrives as mail from that human (the hosted dashboard
sends it as Re: <ref> <title>). It may instead be a task comment, or the task
may be closed. When you are woken, check aw task show <ref> (MCP: task_get
and task_comment_list), read the answer and update the task accordingly.
Claiming work
To take a ready task, mark it in-progress (this is the claim — the team sees you own it):
aw task update <task-id> --status in_progressBefore claiming, run aw work active to make sure nobody is already on the same scope. Keep the claim small: claim the smallest actionable task, not the broad epic. Coordinators may move work around without claiming every subtask.
When status changes, update the task. Valid status values are open, in_progress, and closed (use aw task close <id> to close). Closing with --reason "..." records why; comments capture validation evidence and decisions.
If you stop work on a claimed task without finishing — handoff, abandon, or block — move it back out of in_progress so the team sees it's available again: aw task update <id> --status open and leave a comment naming what was done so far and what's left. Mail the teammate who can unblock or continue.
Wrap-up
Update task status and close finished tasks with
aw task close <id> --reason "what landed and where validated". If work remains, record the handoff or
blocker and release the claim as described above. Follow aweb-coordination
for mail handoffs, locks, and waiting chats.
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
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.

