4n1d

name

4n1d

description

Roadmap triage across the 4N1D framework (NOW/NEXT/NOT-YET/NEVER/DONE in docs/roadmap/). Scans session for completed work, killed ideas, new discoveries, fired triggers, and stale items. Recommends additions, promotions, demotions, and archiving across all five files. Checks anti-patterns (NOW >4, no [learn], stale triggers, zombie NOT-YET). Use at session end, after shipping, or on-demand. Triggers on "/4n1d", "update roadmap", "roadmap triage", "what goes in NOW", "promote to NEXT", "session wrap-up", "what did we ship", or any request to review docs/roadmap/ files.

4N1D Roadmap Triage

End-of-session review that keeps the 4N1D roadmap honest. Read all five files, scan the conversation, recommend changes.

What is 4N1D?

A lightweight roadmap framework built around five files that each answer one question:

File	Question it answers
NOW.md	What are we actively working on? (≤4 items)
NEXT.md	What's queued with a written trigger for promotion?
NOT-YET.md	What ideas are parked until conditions change?
NEVER.md	What have we structurally killed and why?
DONE.md	What did we ship, and what did we learn?

The framework forces every idea to live in exactly one file and forces every promotion to be justified by a written trigger — not vibes.

Process

1. Read current state

Read all files in docs/roadmap/ in parallel:

• NOW.md — current priorities
• NEXT.md — triggered queue
• NOT-YET.md — parked ideas
• NEVER.md — structural kills
• DONE.md — shipped archive
• README.md — framework rules (only if unfamiliar with 4N1D)

2. Scan conversation for signals

Review the current session for:

Signal	Maps to
Completed work (merged PRs, shipped features, closed bugs)	→ DONE.md entries
Killed ideas (adversarial review, infeasible, philosophy violation)	→ NEVER.md (if structural) or NOT-YET.md (if timing)
New ideas discussed but not built	→ NOT-YET.md (default) or NOW.md (if metric-justified)
Evidence that a NEXT trigger fired	→ Promote NEXT → NOW
Evidence that a NOT-YET trigger emerged	→ Promote NOT-YET → NEXT
NOW items worked on but not finished	→ Keep in NOW, note progress
NOW items not touched this session	→ Flag for staleness check

3. Run anti-pattern checks

Flag violations before recommending changes:

• NOW >4 items — recommend cuts
• NOW has no [learn] items — flag "executing without validating"
• NOW item >14 days old — flag for re-justification
• NOW item >30 days old — recommend break-up or demotion
• NEXT triggers >30 days old — flag for refresh
• NOT-YET items >90 days without activity — recommend NEVER or delete
• NEVER >10 items — check for timing-kills that belong in NOT-YET
• DONE.md missing entries for completed work — add them

4. Generate recommendations

Output a structured triage report:

## 4N1D Triage — [date]

### DONE.md (add)
- [type] Description (duration) → outcome

### NOW.md (no changes / add / remove)
- Swap: remove [X], add [Y] because [reason]
- Flag: [item] has been in NOW for [N] days

### NEXT.md (no changes / promote / add / refresh triggers)
- Promote: [item] → NOW (trigger fired: [evidence])
- Refresh: [item] trigger is 45 days old

### NOT-YET.md (no changes / add / promote / delete)
- Add: [new idea from session] with thesis + trigger + kill condition
- Promote: [item] → NEXT (trigger checklist can now be written)

### NEVER.md (no changes / add)
- Add: [idea] — structural kill because [reason]

### Anti-pattern flags
- [list any violations found]

5. Apply changes

After presenting the triage report, ask: "Apply these changes?" If confirmed:

• Edit each file directly (use Edit tool, not full rewrites)
• Add <!-- reviewed: YYYY-MM-DD --> datestamp to each modified file
• For DONE.md: append to the current week's section
• For NEXT.md: include **Trigger written:** YYYY-MM-DD on new items
• For NOW.md: include **Added:** YYYY-MM-DD on new items

Work-type tags

Every item must have one: [build], [learn], [maintain], [decide], [hotfix].

Typed triggers (NEXT items)

Pick the type that matches the decision structure:

Type	Format
Metric-gated	>N [metric]
Qualitative-gated	[signal] from [named verifiable source]
Time-gated	by YYYY-MM-DD because [external constraint]
Dependency-gated	when [item X] reaches DONE
Compound	metric AND qualitative (most common)

NEVER rules

Only structural kills: legal, philosophical, architectural. If the reason is "wrong timing" or "not a priority," it goes in NOT-YET. Keep NEVER under 10 items.

DONE format

One-line entries grouped by ISO week. Include outcome for [learn] and [decide]:

- [build] Feature name (duration)
- [learn] Research topic → key finding
- [decide] Decision made → chosen option + rationale
- [hotfix] What broke — root cause

Bootstrap: if docs/roadmap/ doesn't exist

If the project has no docs/roadmap/ directory yet, offer to bootstrap one. Create five empty files with the section headers below and a short README.md explaining the framework.

docs/roadmap/
├── README.md      # "4N1D — one file per question. See each file's header."
├── NOW.md         # "# NOW\n\nActive work (≤4 items).\n"
├── NEXT.md        # "# NEXT\n\nTriggered queue.\n"
├── NOT-YET.md     # "# NOT-YET\n\nParked ideas.\n"
├── NEVER.md       # "# NEVER\n\nStructural kills.\n"
└── DONE.md        # "# DONE\n\n## Week of YYYY-MM-DD\n"