Suggest candidate epics from a Product Vision, current system state, active epics, and archived epics without creating files until the user accepts a specific suggestion. Use when the user wants to brainstorm, slice a vision, find Product Vision coverage gaps, or generate new epic options.
Scanned 9/4/2026
Install to Claude Code
npx -y skills add NeverSight/skills_feed --skill suggest-epics --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Suggest Epics?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/neversight-suggest-epics)More formats (shields.io, HTML) on the badges page.
---
name: suggest-epics
description: Suggest candidate epics from a Product Vision, current system state, active epics, and archived epics without creating files until the user accepts a specific suggestion. Use when the user wants to brainstorm, slice a vision, find Product Vision coverage gaps, or generate new epic options.
---
# Suggest Epics
## Core Rule
Suggest candidate epic boundaries from the single project Product Vision; do not create or modify epic files unless the user explicitly accepts one or more suggestions and asks you to proceed with `create-epic`. This skill separates Product Vision decomposition from artifact creation. Follow [references/clarification-policy.md](references/clarification-policy.md).
## Workflow
1. Load inputs:
- Read the source Product Vision from `docs/vision.md`.
- Read existing `docs/epics/active/*/epic.md`.
- Read relevant `docs/epics/archived/*/epic.md` and archive summaries when they inform overlap, precedent, cancelled scope, deferred scope, sequencing, or lessons from completed work.
- Read relevant `docs/knowledge/` concern specs and inspect code or docs enough to understand current system boundaries.
2. Analyze coverage:
- Map current Product Vision goals, success signals, principles, constraints, non-goals, assumptions, and open questions to active and relevant archived epics.
- Identify already-owned scope, partial coverage, gaps, conflicts, duplicated ownership, and sequencing constraints.
- Use archived epics to avoid re-suggesting completed, cancelled, superseded, or intentionally deferred work without calling out why it is being reopened.
- Distinguish product or platform outcomes from technical layers.
- Treat full Product Vision coverage analysis as suggestion input, not as epic validation.
3. Suggest candidates:
- Propose the smallest useful set of candidate epics that cover missing or requested Product Vision outcomes.
- For each candidate, include outcome, why it is end-to-end, key in-scope and out-of-scope items, Product Vision coverage, dependencies, sequencing, overlap risk, and open questions.
- Prefer fewer, clearer candidates. Do not force a full breakdown if the user asked about one area.
4. Ask for acceptance:
- Ask the user which candidate to create or revise.
- If accepted, use `create-epic` for each accepted candidate, one epic at a time.
- If rejected or changed, revise suggestions before writing any artifact.
## Validation Checklist
- [ ] Existing active epics were checked for coverage and overlap.
- [ ] Relevant archived epics and archive summaries were checked for precedent, completed scope, cancelled scope, deferred scope, and lessons learned.
- [ ] Suggestions are end-to-end delivery outcomes, not technical layers.
- [ ] Each suggestion maps to specific Product Vision items.
- [ ] Gaps, overlap, dependencies, sequencing, and open questions are explicit.
- [ ] No files were created or modified unless the user accepted a suggestion and requested creation.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!