Agent skill
create-roadmap
Create a Now/Next/Later outcome-oriented roadmap when the user asks to plan a roadmap, prioritize features, or define product direction
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/create-roadmap
SKILL.md
Create Roadmap
Overview
Generate a Now/Next/Later roadmap where every item ties to a measurable outcome, not just a feature name. This format communicates strategic intent without false precision on dates, and forces prioritization by time horizon and confidence level.
Workflow
-
Read product context -- Scan
.chalk/docs/product/for the product profile, PRDs, JTBD docs, and metrics frameworks. Check.chalk/docs/engineering/for architecture constraints and tech debt items. These inform what is feasible and what matters. -
Parse scope -- Extract from
$ARGUMENTSthe product area, theme, or time horizon to roadmap. If unspecified, create a full-product roadmap using all available context. -
Determine the next file number -- Read filenames in
.chalk/docs/product/to find the highest numbered file. The next number ishighest + 1. -
Categorize items into time horizons:
- Now (current cycle, high confidence): Actively being built or about to start. Clear problem definition, validated need, committed resources.
- Next (1-2 cycles out, medium confidence): Problem is understood, solution is taking shape, not yet committed.
- Later (future, low confidence): Strategic bets, exploration areas, things that depend on learning from Now/Next.
-
Write each item with columns: Item, Outcome (measurable result), Confidence (high/medium/low), Dependencies (what must be true or done first).
-
Add strategic context -- Above the roadmap table, write a brief narrative explaining the strategic themes connecting the items and what the roadmap is optimizing for.
-
Write the file -- Save to
.chalk/docs/product/<n>_roadmap.md. -
Confirm -- Share the file path and highlight any items with unresolved dependencies or low confidence that need validation.
Output
- File:
.chalk/docs/product/<n>_roadmap.md - Format: Plain markdown with strategic narrative and Now/Next/Later tables
- First line:
# Roadmap: <Product Area or Theme>
Anti-patterns
- Feature lists without outcomes -- "Add filtering" is not a roadmap item. "Users can find records in <10 seconds (reduce support tickets by 30%)" ties the work to value.
- Dates on everything -- Now/Next/Later intentionally avoids specific dates for Next and Later items. Adding dates to low-confidence items creates false promises.
- No dependencies called out -- Roadmap items without dependencies hide sequencing risks. If B requires A, say so.
- Confidence theater -- Marking everything as "high confidence" defeats the purpose. Be honest about what is validated versus what is a bet.
- Solution-shaped items -- "Build GraphQL API" belongs in an engineering plan, not a product roadmap. Roadmap items should describe user or business outcomes.
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?