메뉴
Decompose a PRD into Epics and INVEST-shaped User Stories with acceptance criteria and priority. Invoke when the team has a baselined PRD and needs an actionable story backlog before sprint planning or WBS construction.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Route the user to the right brainstorming strategy for their situation by matching their goal to a curated shortcut table or, when the goal is unclear, asking 3 diagnostic questions and recommending one or two strategies from the curated 20. Invoke when the user wants to brainstorm but has not yet chosen a technique.
Facilitate a Scrum Sprint Planning event, producing a sprint goal and a committed Sprint Backlog drawn from the prioritized Product Backlog. Invoke when the team is about to begin a new sprint and needs to select and plan its work.
Construct a PMBOK-shaped Work Breakdown Structure from the project's epic-and-story hierarchy. Invoke when the team needs a deliverable-oriented decomposition of project scope into manageable work packages before building a schedule or assigning resources.
Draft a PMBOK-shaped project charter from a concept memo or product brief. Invoke when the user needs a charter for initiation-phase documentation.
Produce a PMBOK-shaped project closure report suitable for sponsor signature, consolidating charter, final schedule, final budget, retrospectives, post-mortems, and stakeholder acceptance records. Invoke when the project has reached final deliverable acceptance and requires formal closure documentation.
Produce a PMBOK-shaped communication matrix, escalation paths, and meeting cadence from a project charter and stakeholder register. Invoke when the team needs a formal plan governing how project information flows to each audience.
| name | epics-and-stories |
| description | Decompose a PRD into Epics and INVEST-shaped User Stories with acceptance criteria and priority. Invoke when the team has a baselined PRD and needs an actionable story backlog before sprint planning or WBS construction. |
Adapted from bmad-method:bmad-create-epics-and-stories (MIT, © 2025 BMad Code, LLC). See THIRD_PARTY_NOTICES.md.
Use this skill after the PRD has been reviewed and baselined. The skill is the bridge between a requirements document and an executable sprint backlog. Typical triggers include:
Do not invoke this skill to produce the PRD itself (use the prd skill). Do not use it in place of the WBS — epics and stories decompose scope by feature value; the WBS decomposes scope by deliverable. Both views are complementary, not substitutes.
Before facilitating story decomposition, load the canonical User Story reference to ground acceptance criteria and the INVEST framework in established practice.
Reading the config. Check .pm-kit.config.json for the sourcesMode field:
sourcesMode is "online" (opt-in): fetch the URL stored at key sources.epicsAndStories in vendor/pm-kit/sources-index.json using your available web-fetch capability. Do not name a specific tool — use whatever your runtime provides. Ground the facilitation in what you read. Do not fabricate quotations or page numbers.sourcesMode is "offline" or the field is absent (the default): rely on your general knowledge of the Agile user story format (As a / I want / So that), the INVEST criteria (Independent, Negotiable, Valuable, Estimable, Small, Testable), and Gherkin-style acceptance criteria (Given / When / Then) as described by the Agile Alliance and widely adopted Scrum practice. Cite the canonical URL from vendor/pm-kit/sources-index.json at key sources.epicsAndStories in the output. Do not fabricate quotations or page numbers.In both cases, the URL to cite is https://www.agilealliance.org/glossary/user-story.
Walk the user through these steps in sequence. Do not skip steps or combine them.
Step 1 — Ingest inputs. Ask the user to supply (or confirm the location of) the baselined PRD. Identify the list of functional requirements (FR-xx items). Confirm the inputs are present before proceeding.
Step 2 — Group requirements into Epics. For each cluster of closely related functional requirements, propose an Epic: a large, coherent feature grouping that delivers visible value to a user type. Present a draft Epic list to the user. Revise until the user confirms the groupings.
Step 3 — Write User Stories per Epic. For each Epic, decompose it into atomic User Stories. Each story must follow the format: "As a <user type>, I want <action or capability>, so that <outcome or value>." Stories must satisfy INVEST — if a story is not independently deliverable or not estimable, split or merge it before proceeding.
Step 4 — Write acceptance criteria. For each story, elicit at least two acceptance criteria using Given / When / Then notation. Each criterion must be independently testable by the acceptance team without additional clarification.
Step 5 — Assign priority. For each story, assign a priority of Must (delivers core value; release is not viable without it), Should (high value; include if possible), or Could (nice to have; defer if constrained). Confirm the priority distribution with the user.
Step 6 — Tag dependencies. Identify any story that depends on another story being completed first. Record the dependency by story ID.
Step 7 — Summarize the backlog. Produce the Backlog Summary section: total epics, total stories, aggregate story-point placeholder (to be populated by the cost-estimation skill), and priority distribution.
Step 8 — Save the artifact. Save the filled artifact to docs/pm-kit/outputs/epics-and-stories/<short-slug>.md. <short-slug> is a kebab-case ASCII slug (max 40 chars) derived from the project name (e.g., bookswap-campus). Confirm the final path with the user before writing. If the file exists, ask: overwrite, append a date suffix (e.g., -2026-04-20), or pick a new slug. The artifact begins with the three-line provenance header below (HTML comments, do not render):
<!-- Generated by agentic-pm-kit:epics-and-stories on YYYY-MM-DD -->
<!-- Languages: communication=<value>, output=<value> -->
<!-- Source mode: offline | online -->
The kit separates the language used for live agent–user dialogue from the language used in the saved artifact. Both values live in .pm-kit.config.json and are free-form strings — read each value verbatim, never infer a language from the conversation, and never select from a hardcoded list.
Facilitation dialogue. Speak to the user during facilitation in the language at language.communication. Use the string verbatim.
Filled artifact (saved TEMPLATE.md output). Produce the written artifact in the language at language.output. If language.output is absent or empty, fall back to language.communication.
Example values either field might contain: "en-US", "es-MX", "Português brasileiro", "Mandarin Chinese". Accept any string as given.
When the backlog is complete, point the user to CHECKLIST.md (sibling file) and ask them to verify each item. Remind them that the output must be marked PASS or FAIL. On FAIL, invite the user to return with specific notes so the facilitation can be resumed or corrected.