|
| 1 | +# Spec of Specs |
| 2 | + |
| 3 | +When a feature is too large to run through a single |
| 4 | +`/speckit.specify` → `/speckit.plan` → `/speckit.tasks` → `/speckit.implement` |
| 5 | +cycle without the model losing track mid-implementation, you can break it into a |
| 6 | +**roadmap** of smaller, independently-specified sub-features. This is the "spec of |
| 7 | +specs" approach: one up-front pass decomposes a massive feature into self-contained |
| 8 | +specs, and each of those runs through its own specify/plan/tasks/implement cycle. |
| 9 | + |
| 10 | +> **When to reach for this.** Decomposition adds the most overhead of any strategy |
| 11 | +> in [Handling Complex Features](complex-features.md). Use it **only when the lighter |
| 12 | +> options there are insufficient** — first try limiting how many tasks run per |
| 13 | +> `/speckit.implement` invocation, then sub-agent delegation, then a combination. |
| 14 | +> Reach for a spec of specs only when even a single phase is too large to handle in |
| 15 | +> one run. |
| 16 | +
|
| 17 | +The rest of this page describes *how* to do it with the tools you already have. No |
| 18 | +new commands or extensions are required. |
| 19 | + |
| 20 | +## The roadmap pass |
| 21 | + |
| 22 | +Before writing any sub-spec, do a single decomposition pass to produce a roadmap. |
| 23 | +Treat this as a lightweight planning conversation with your agent, not a full spec: |
| 24 | + |
| 25 | +1. **State the whole feature.** Describe the large feature (the "epic") in a |
| 26 | + sentence or two so the agent has the full picture up front. |
| 27 | +2. **Identify independent slices.** Ask the agent to propose a small set of |
| 28 | + sub-features that each deliver a coherent piece of the epic and can be specified |
| 29 | + on their own. Aim for slices that are independently testable — implementing just |
| 30 | + one should leave you with something demonstrable. |
| 31 | +3. **Draw the boundaries.** For each slice, write one line of intent and an explicit |
| 32 | + scope boundary (what is in, what is deferred to a sibling slice). Sharp |
| 33 | + boundaries are what keep each sub-spec small enough to fit in context. |
| 34 | +4. **Order by dependency.** Note which slices depend on others and sequence them so |
| 35 | + prerequisites come first. Slices with no dependency on each other can be built in |
| 36 | + any order (or in parallel). |
| 37 | +5. **Record the result as a roadmap.** Capture the slices in a durable roadmap file |
| 38 | + (below) so every later sub-spec can point back to it. |
| 39 | + |
| 40 | +The roadmap is deliberately shallow: it names and orders the sub-features but does |
| 41 | +**not** design them. The design happens when each slice runs through its own |
| 42 | +`/speckit.specify`. |
| 43 | + |
| 44 | +## The roadmap artifact |
| 45 | + |
| 46 | +The roadmap is an ordinary Markdown file you author and keep under version control — |
| 47 | +there is no special tooling behind it. Put it where the sub-specs can find it: |
| 48 | + |
| 49 | +- For a feature-scoped epic: `specs/<epic-slug>/roadmap.md`. |
| 50 | +- For a larger, cross-cutting epic: a top-level `ROADMAP.md`. |
| 51 | + |
| 52 | +Each roadmap entry carries a stable id (used later for linking), a name, its intent, |
| 53 | +its scope boundary, its dependencies, a status, and — once the sub-spec exists — a |
| 54 | +link to it. A minimal template: |
| 55 | + |
| 56 | +```markdown |
| 57 | +# Roadmap: <epic name> |
| 58 | + |
| 59 | +<One or two sentences: what the epic is and why it is being decomposed.> |
| 60 | + |
| 61 | +**Status legend**: planned · in-progress · done |
| 62 | + |
| 63 | +| ID | Sub-feature | Intent | Scope boundary | Depends on | Status | Sub-spec | |
| 64 | +|----|-------------|--------|----------------|-----------|--------|----------| |
| 65 | +| R1 | <name> | <one line> | <in / deferred> | — | planned | — | |
| 66 | +| R2 | <name> | <one line> | <in / deferred> | R1 | planned | — | |
| 67 | +| R3 | <name> | <one line> | <in / deferred> | R1 | planned | — | |
| 68 | +``` |
| 69 | + |
| 70 | +Keep the `ID` column immutable once a sub-spec references it — it is the anchor for |
| 71 | +traceability. Fill in the `Sub-spec` column with the path to each sub-feature's spec |
| 72 | +directory as you create it, and update `Status` as work progresses. |
| 73 | + |
| 74 | +## Specifying each sub-feature |
| 75 | + |
| 76 | +With the roadmap in hand, work through the entries one at a time using the normal |
| 77 | +Spec Kit flow — nothing new to learn: |
| 78 | + |
| 79 | +1. Pick the next roadmap entry whose dependencies are already `done` (or have none). |
| 80 | +2. Run `/speckit.specify` for just that slice, describing only its intent and scope |
| 81 | + from the roadmap entry. Because the slice is bounded, its spec, plan, and tasks |
| 82 | + stay well within the context window. |
| 83 | +3. Run `/speckit.plan`, `/speckit.tasks`, and `/speckit.implement` for that slice as |
| 84 | + usual. |
| 85 | +4. Mark the roadmap entry `done` and move to the next one. |
| 86 | + |
| 87 | +Each slice is a complete, independent Spec Kit feature with its own |
| 88 | +`spec.md`/`plan.md`/`tasks.md`. The roadmap is what ties them together. |
| 89 | + |
| 90 | +## Linking sub-specs to the roadmap |
| 91 | + |
| 92 | +To keep scope and intent from drifting across separate runs, every sub-spec |
| 93 | +references its roadmap entry, and the roadmap links back — a simple, greppable, |
| 94 | +bidirectional convention: |
| 95 | + |
| 96 | +- **Sub-spec → roadmap.** In the sub-feature's `spec.md`, name the parent roadmap |
| 97 | + and entry id in the `Input` / summary line, for example: |
| 98 | + |
| 99 | + ```markdown |
| 100 | + **Input**: Parent roadmap: `specs/<epic>/roadmap.md` → entry **R3**. <feature description> |
| 101 | + ``` |
| 102 | + |
| 103 | +- **Roadmap → sub-spec.** In the roadmap table, set the entry's `Sub-spec` column to |
| 104 | + the sub-feature's directory, e.g. `specs/<epic>-part-3/`. |
| 105 | + |
| 106 | +Because both directions are plain text, you can trace any sub-spec back to its place |
| 107 | +in the epic (and find its siblings) with a quick search — no tooling, no metadata |
| 108 | +schema. |
| 109 | + |
| 110 | +## Keeping the roadmap and sub-specs in sync |
| 111 | + |
| 112 | +The roadmap is a living document. As you learn more, keep it and the sub-specs |
| 113 | +aligned: |
| 114 | + |
| 115 | +- **Roadmap first, then reconcile.** When scope shifts, update the roadmap entry |
| 116 | + first, then update any sub-specs it affects. The roadmap is the source of truth for |
| 117 | + how the epic is divided. |
| 118 | +- **Respect dependencies and ordering.** If a slice depends on another, build the |
| 119 | + prerequisite first and cross-reference the dependent sub-spec so the relationship |
| 120 | + is visible from both sides. |
| 121 | +- **Recurse when a slice is still too big.** If a sub-feature turns out to be too |
| 122 | + large to specify in one cycle, give it its own roadmap and decompose it further — |
| 123 | + the same approach applies one level down. Recursion adds overhead, so only go as |
| 124 | + deep as the context problem actually requires. |
| 125 | + |
| 126 | +## Worked example |
| 127 | + |
| 128 | +Suppose the epic is **"Add a self-service billing portal"** — far too large for a |
| 129 | +single cycle. The roadmap pass breaks it into three independently-specifiable |
| 130 | +slices. |
| 131 | + |
| 132 | +`specs/billing-portal/roadmap.md`: |
| 133 | + |
| 134 | +```markdown |
| 135 | +# Roadmap: Self-service billing portal |
| 136 | + |
| 137 | +Let customers view invoices, manage payment methods, and change plans without |
| 138 | +contacting support. Too large for one cycle, so it is split into independent slices. |
| 139 | + |
| 140 | +**Status legend**: planned · in-progress · done |
| 141 | + |
| 142 | +| ID | Sub-feature | Intent | Scope boundary | Depends on | Status | Sub-spec | |
| 143 | +|----|--------------------|------------------------------------------|---------------------------------------------|-----------|---------|----------| |
| 144 | +| R1 | Invoice history | Customers view and download past invoices | Read-only; no payment actions | — | done | specs/billing-invoices/ | |
| 145 | +| R2 | Payment methods | Add, remove, and set a default card | No plan changes; assumes invoices exist | R1 | in-progress | specs/billing-payment-methods/ | |
| 146 | +| R3 | Plan changes | Upgrade/downgrade the subscription plan | Uses R2's default payment method | R1, R2 | planned | — | |
| 147 | +``` |
| 148 | + |
| 149 | +Each slice is then specified on its own. For example, the **R2** sub-feature's |
| 150 | +`spec.md` opens with a back-reference: |
| 151 | + |
| 152 | +```markdown |
| 153 | +# Feature Specification: Billing — payment methods |
| 154 | + |
| 155 | +**Input**: Parent roadmap: `specs/billing-portal/roadmap.md` → entry **R2**. |
| 156 | +Let customers add, remove, and set a default payment method in the billing portal. |
| 157 | +``` |
| 158 | + |
| 159 | +From here a reader can trace **R2** back to the roadmap, see that it depends on |
| 160 | +**R1** (invoice history, already `done`), and see that **R3** (plan changes) is |
| 161 | +waiting on it. Building R1, then R2, then R3 keeps every run small while the roadmap |
| 162 | +preserves the shape of the whole epic. |
| 163 | + |
| 164 | +## For automation (optional) |
| 165 | + |
| 166 | +If you would rather automate roadmap capture and consistency checks than maintain |
| 167 | +the file by hand, the community-maintained |
| 168 | +[Spec Roadmap extension](https://github.com/srobroek/speckit-roadmap) explores that |
| 169 | +direction. It is a third-party extension and is not required — the manual convention |
| 170 | +above is enough on its own. |
0 commit comments