ctx-next

name	ctx-next
description	Suggest what to work on next. Use when starting a session, finishing a task, or when unsure what to prioritize.
allowed-tools	Bash(ctx:*), Read

Analyze current tasks and recent session activity, then suggest 1-3 concrete next actions with rationale.

When to Use

• At session start after loading context ("what should I do?")
• After completing a task ("what's next?")
• When the user asks for priorities or direction
• When multiple tasks exist and it's unclear which to pick

When NOT to Use

• When the user has already stated what they want to work on
• When actively mid-task (don't interrupt flow with suggestions)
• When no context directory exists (nothing to analyze)

Usage Examples

/ctx-next
/ctx-next (just finished the auth refactor)

Process

Do all of this silently: do not narrate the steps:

1. Read TASKS.md to get the full task list with statuses, priorities, and phases
2. Check recent sessions to understand what was just worked on and avoid suggesting already-completed work:

   ctx journal source --limit 3
   

3. Read the most recent session file (if any) to understand what was accomplished and what follow-up items were noted
4. Analyze and rank tasks using the priority logic below
5. Present 1-3 recommendations in the output format below

Priority Logic

Rank candidate tasks using these criteria (in order):

1. Explicit priority: #priority:high > #priority:medium

   #priority:low > untagged

2. Unblocked: tasks not tagged #blocked or listed under a "Blocked" section
3. In-progress first: #in-progress tasks should be resumed before starting new ones (finishing > starting)
4. Momentum: prefer tasks related to recent session work (continuing a thread is cheaper than context-switching)
5. Phase order: earlier phases before later phases (Phase 0 before Phase 1, etc.) unless priority overrides
6. Quick wins: if two tasks have equal priority, prefer the one that seems smaller/faster (builds momentum)

Skip these tasks:

• [x] completed tasks
• [-] skipped tasks
• Tasks explicitly tagged #blocked with no resolution path
• Tasks that were the main focus of the most recent session (user likely wants variety or the session ended because it was done)

Output Format

Present your recommendations like this:

Recommended Next

1. [Task title or summary] #priority:X

[1-2 sentence rationale: why this, why now]

2. [Task title or summary] #priority:X

[1-2 sentence rationale]

3. [Task title or summary] (optional: only if genuinely useful)

[1-2 sentence rationale]

---

Based on N pending tasks across M phases. Last session: [topic] ([date]).

Rules for recommendations:

• 1-3 items only: more than 3 defeats the purpose
• Be specific: "Fix block-non-path-ctx hook" not "work on hooks"
• Include the priority tag so the user sees the weight
• Rationale must reference context: why this task, not just what it is. Connect to recent work, priority, or dependencies
• If an in-progress task exists, it should almost always be recommendation #1 (don't abandon unfinished work)

Examples

Good Output

Recommended Next

1. Fix block-non-path-ctx hook #priority:high

Still open from yesterday's session. The hook is too aggressive: it blocks git -C path commands that don't invoke ctx. Quick fix, clears a blocker.

2. Add Context.File(name) method #priority:high

Eliminates 10+ linear scan boilerplate instances across 5 packages. High impact, low effort: good consolidation target.

3. Topics system (T1.1) #priority:medium

Journal site's most impactful remaining feature. Metadata is already in place from the enrichment work.

---

Based on 24 pending tasks across 3 phases. Last session: doc-drift-cleanup (2026-02-11).

Bad Output

"You have many tasks. Here are some options:

• Do some stuff with hooks
• Maybe work on tests
• There's also some docs to write"

(Too vague, no priorities, no rationale, no connection to context.)

Quality Checklist

Before presenting recommendations, verify:

• TASKS.md was read (not guessed from memory)
• Recent sessions were checked to avoid re-suggesting completed work
• Each recommendation has a specific task reference
• Each recommendation has a rationale grounded in context
• In-progress tasks are prioritized over new starts
• No more than 3 recommendations
• Footer shows task count and last session reference