agentleFS
MicrosoftKnown publishermicrosoft/devsquad-copilot/.github/plugins/devsquad/skills/next-task/SKILL.md

next-task

Skill50 starsChanged 16 days agoMITView on GitHub

Next task suggestion after completing implementation. Use when the developer finishes a task and wants a recommendation for the next one. Do not use during an ongoing implementation, for initial task selection (use work-item-workflow), or for sprint planning (use devsquad.sprint).

We split this skill into the rules it gives an agent

12 rules. 0 are common, 9 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After completing implementation (PR created or push performed):

    line 10
  2. Distinctive:

    Before presenting suggestions, check if there are open PRs in the repository that haven't been merged yet.

    line 36
  3. Distinctive:

    If there are open PRs, check if the suggested tasks depend on the code in those PRs (same user story, later phase, or explicit dependency …

    line 44
  4. Distinctive:

    If the selected task depends on code from an open PR (not merged), alert:

    line 68
  5. Distinctive:

    If the user selects a task:

    line 84
  6. Distinctive:

    If there are uncommitted changes:

    line 93
  7. Distinctive:

    If there are unpushed commits:

    line 105
  8. Very specific:

    If the current branch is not main/master/develop:

    line 118
  9. Distinctive:

    After ensuring a clean branch:

    line 131
  10. Very specific:

    Read branching strategy: cat .memory/git-config.md 2>/dev/null

    line 133
  11. Very specific:

    If it doesn't exist, use fallback: git symbolic-ref refs/remotes/origin/HEAD | sed 's@^refs/remotes/origin/@@'

    line 134
  12. Distinctive:

    Create new branch for the selected task

    line 137

“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Next Task — Next Task Suggestion

Offer Suggestion

After completing implementation (PR created or push performed):

Task completed! Would you like a suggestion for the next task?

[Y] Yes, suggest next task
[N] No, end session

Find Available Tasks

  • GitHub: Issues with label type:task, no assignee, state open
  • Azure DevOps: Work items of type Task, no Assigned To, state New/To Do, in the current sprint

Sort by Criteria

CriterionWeightDescription
PriorityHighP1 > P2 > P3
Dependencies satisfiedHighTasks without blockers first
Same User StoryMediumContext continuity
Same featureMediumDomain continuity
Sequential phaseLowSetup → Foundation → Core → Integration

Check Open PRs

Before presenting suggestions, check if there are open PRs in the repository that haven't been merged yet. The dependency is about code that is not in main, regardless of who opened the PR.

# List open PRs in the repository
# GitHub: github/list_pull_requests(owner, repo, state: "open")
# Azure DevOps: ado/repo_pull_request(action: "list", status: "active")

If there are open PRs, check if the suggested tasks depend on the code in those PRs (same user story, later phase, or explicit dependency in tasks.md).

Present Suggestions (top 3)

Recommended next tasks:

1. #[number]: [title]
   Priority: P1 | Feature: [name] | Phase: [phase]
   Reason: Highest priority, same feature

2. #[number]: [title]
   Priority: P1 | Feature: [name] | Phase: [phase]
   Reason: Dependency from previous task satisfied

3. #[number]: [title]
   Priority: P2 | Feature: [name] | Phase: [phase]
   Reason: User Story continuity

[1/2/3] Select task
[O] See other options
[N] End session

If the selected task depends on code from an open PR (not merged), alert:

Warning: this task depends on code from PR #[number] ([title]), which has not been merged yet.

If you create the branch from the main branch, the code from the previous task will not be available.

[W] Wait for PR merge and continue later
[B] Create branch from the previous branch ([branch-name]) — stacked branch
[I] Ignore and create from the main branch anyway

Note: Option [B] (stacked branch) works, but requires rebase after the previous PR is merged. Alert the dev about this implication.

Branch Transition

If the user selects a task:

Check pending changes

git status
git log origin/$(git branch --show-current)..HEAD --oneline 2>/dev/null

If there are uncommitted changes:

You have uncommitted changes on the current branch:

[modified files]

[C] Commit changes and continue
[S] Stash and continue (not recommended)
[A] Abort and continue on the current task

If there are unpushed commits:

You have local commits not pushed to remote:

[list of commits]

[P] Push and continue
[A] Abort and continue on the current task

Return to main branch

If the current branch is not main/master/develop:

You are on branch: [branch-name]

To keep branches short-lived, I recommend:

[M] Return to main/develop and create a new branch
[C] Continue on the current branch (not recommended for different tasks)

Start new task

After ensuring a clean branch:

  • Read branching strategy: cat .memory/git-config.md 2>/dev/null
  • If it doesn't exist, use fallback: git symbolic-ref refs/remotes/origin/HEAD | sed 's@^refs/remotes/origin/@@'
  • Return to integration branch: git checkout <integration-branch>
  • Update: git pull origin <integration-branch>
  • Create new branch for the selected task
  • Start implementation flow
  • Apply all checks (assignee, dependencies, priority, capacity)

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.