Agent skill
next
Pick the next bead to work on. Shows ready tasks (no blockers), applies user preferences for ordering (priority, type, recency), and helps select work.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/next-flurdy-agent-skills
SKILL.md
Next - Pick Your Next Bead
Help select the next bead to work on based on readiness and user preferences.
When to Use
- Starting a new work session
- Finished a task and need to pick the next one
- Want to see what's available to work on
- Need help prioritizing between multiple options
Usage
/next # Show ready beads, ranked by suitability
/next task # Auto-pick the next most suitable task and start it
/next quick # Auto-pick an easy win task and start it
/next bug # Auto-pick the next most important bug and fix it
/next <bead-id> # Start working on specific bead
What This Skill Does
-
Find Ready Work
- Run
bd list --readyto get open, unblocked tasks - Excludes
in_progressbeads (another session may be working on them) - Show current in-progress work if any (for awareness, not selection)
- Run
-
Rank by Suitability
- Apply priority ranking algorithm (see below)
- Bugs generally rank higher than features at same priority
- Epics rank lower (they represent larger work)
-
Present Options
- Show top 5 candidates with key details
- Include: ID, title, priority, type, age
- Ask user to pick or provide different criteria
-
Start Work
- Mark selected bead as in_progress
- Show full bead details
- Suggest first steps if description includes them
Examples
# Show ready work ranked by suitability
/next
# Auto-pick and start the next most suitable task
/next task
# Auto-pick an easy win (quick task)
/next quick
# Auto-pick the next most important bug and start fixing
/next bug
# Start a specific bead
/next gauge-abc
Output Format
## Ready to Work (5 of 12 open)
| # | ID | Pri | Type | Parent/Subs | Title |
|---|-----------|-----|---------|-------------|--------------------------------|
| 1 | gauge-abc | P1 | bug | - | Fix login timeout issue |
| 2 | gauge-def | P2 | feature | 3 subtasks | Add export to CSV |
| 3 | gauge-ghi | P2 | task | gauge-def | Update dependencies |
| 4 | gauge-jkl | P3 | feature | - | Dark mode toggle |
| 5 | gauge-mno | P3 | task | 2 subtasks | Refactor auth service |
Currently in progress: gauge-xyz "Implement caching layer"
Which would you like to work on? (1-5, or specify ID, or "task" to auto-pick)
Implementation
When invoked:
-
Check for current open, not in-progress elsewhere, work:
bashbd list --status=open -
Get ready (unblocked) beads with open status only, excluding P4 backlog:
bashbd list --ready --priority-max=3Important:
- Use
bd list --ready(notbd ready) to excludein_progressbeads - Use
--priority-max=3to exclude P4 backlog items (P4 = future/someday, never auto-pick) - Another session may be working on in_progress items - picking them up causes conflicts
- Use
-
Parse command argument:
- (none): Show ranked list, ask user to pick
task: Auto-select top-ranked bead and start itquick: Auto-select an easy win task and start itbug: Auto-select top-ranked bug and start it (see Bug Mode below)<bead-id>: Start that specific bead
-
If specific bead ID provided:
bashbd show <id> bd update <id> --status=in_progress -
Otherwise, present top 5 options and ask user to choose
-
On selection:
- Mark as in_progress
- Show full details with
bd show - If bead has description with steps, highlight first step
Handling Edge Cases
- No ready beads (P0-P3): Show blocked beads and what's blocking them; mention P4 backlog exists if any, but don't auto-pick
- All open beads in progress: Warn that another session may be working on them; ask user if they want to see in_progress beads anyway (may cause conflicts)
- User picks in_progress bead: Warn that another session may be working on it; require explicit confirmation before starting
- Invalid ID: Show error and list valid options
- User says "skip": Show next 5 options
Priority Ranking Algorithm
Rank ready beads in this order (first match wins):
| Rank | Criteria |
|---|---|
| 1 | Any P0 issue (any type) |
| 2 | P1 bug |
| 3 | P2 bug |
| 4 | P1 feature or task |
| 5 | P1 epic |
| 6 | P2 feature or task |
| 7 | P3 bug, feature, or task |
| 8 | P2 epic |
| 9 | P3 epic |
| 10 | Any other non-P4 issue |
Important: P4 items are backlog/future work and must NEVER be auto-picked. Always use --priority-max=3 to exclude them. Only show P4 items if user explicitly requests them.
Quick Task Heuristics
When /next quick is used, prefer:
- Type: task > bug > feature (tasks are usually smaller)
- Priority: P3 > P2 > P1 (lower priority = less complex)
- Exclude epics (too large for quick wins)
- Title keywords: "fix", "update", "add" > "implement", "refactor", "redesign"
Bug Mode
When /next bug is used:
-
Filter to open bugs only (excluding P4 backlog):
bashbd list --ready --type=bug --priority-max=3 -
Rank by priority: P0 > P1 > P2 > P3 (highest priority bug first, P4 excluded)
-
Auto-select and start the top-ranked bug
-
Continue fixing bugs if the completed bug was minor:
- After completing a bug fix, assess if it was minor (small change, localized fix)
- If minor AND there's remaining context (related code still fresh), auto-pick the next bug
- Continue this loop until:
- A bug requires significant work (not minor)
- No more ready bugs remain
- Context would be lost (unrelated area of codebase)
Minor Bug Criteria
A bug is considered minor if:
- Fix touches ≤ 3 files
- Change is ≤ 50 lines total
- No architectural changes required
- Fix is localized (single component/module)
Context Continuity
Continue to next bug automatically when:
- Next bug is in same or adjacent files
- Next bug is in same module/component
- Fix for previous bug provides context for next bug
Stop and ask user when:
- Next bug is in completely different area of codebase
- Next bug appears complex (P0/P1 with unclear scope)
- 3+ bugs have been fixed in sequence (natural checkpoint)
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?