Software Development
Spec-First Enforcer
Produces a structured specification with problem definition, inputs, outputs, constraints, and success criteria before any implementation begins. Useful for preventing semantic drift and aligning on scope before writing code. Engineers working on non-trivial features, tech leads who want to enforce design-before-code discipline, developers working with AI agents and tired of re-doing work because the agent solved the wrong problem. the implementation solves a slightly different problem than the one the developer had in mind. The fix is not better AI; it is a clear spec before code is written. Teams skip this step because writing a spec feels slower than writing code — but a 10-minute spec saves a 2-hour retrofit. A structured enforcer turns "write the spec" into a fast, templated exercise that produces a reviewable artifact and a go/no-go readiness score.
One-Time Purchase
$19.99
Spec — User Notification Preferences
Size: Small–Medium (half to one day of work) · Readiness: Conditional — 2 open questions blocking
Summary
Users currently receive every platform notification with no way to control frequency, channel, or type. The result is alert fatigue and users disabling notifications outright. This feature adds a self-service preference center for in-app and email notifications, with admin-enforced overrides for mandatory categories. Storage and UX only — no sending logic lives in this feature.
Requirements
Inputs
Outputs
Users
Acceptance Criteria
| # | Criterion | Verification | Priority |
|---|---|---|---|
| AC-1 | Preference changes take effect within one event cycle (≤30s) | Integration test: flip preference, send qualifying event, assert respected | P0 |
| AC-2 | Admin-enforced preferences render read-only with explanation text | Visual test + unit test on the toggle component | P0 |
| AC-3 | Preference read adds ≤50ms to page load (cached reads expected) | Perf benchmark in CI | P1 |
| AC-4 | Save completes in ≤400ms at p95 under normal load | Load test at expected concurrency | P1 |
| AC-5 | Concurrent saves do not lose changes (last-write-wins w/ optimistic lock is OK if documented) | Concurrency test + ADR note | P1 |
| AC-6 | Audit-log entry is present for every preference mutation | Integration test asserts log row per mutation | P2 |
Open Questions
OQ-1 — Notification type registry: static or dynamic?
If static (hardcoded), the UI is trivial and ships fast. If dynamic (fetched at runtime), we need a loading state, an error path, and a cache-invalidation story. Two materially different builds. Owner: backend lead. Blocks implementation.
OQ-2 — Admin scope: defaults or enforcement?
Admins might (a) set default preferences for new users, (b) enforce mandatory categories no user can disable, or (c) both. Defaults = pre-population logic; enforcement = lock UI; both = both. Owner: product. Blocks implementation.
Out of Scope
Explicitly not in this spec
- SMS / push notification channels
- Notification scheduling ("only notify me between 9am–5pm")
- Per-item muting (e.g., a single document vs. a category)
- Admin ability to blast-override individual user preferences
- Digest / batching configuration
Each of these is a reasonable future request; none of them are blockers for shipping the v1 surface described above.
Verdict
Conditional approval
6 of 8 sections fully resolved. Implementation must not begin until OQ-1 and OQ-2 are answered. Both are resolvable in a 30-minute sync with product and the backend lead — no additional research required. Once resolved, update Inputs and the admin section under Constraints, then promote this spec to Ready.
This sample illustrates the skill's output format. The feature shape is illustrative.
View full sample →
All sales final. No refunds on digital products.
Includes support for Claude Code, Codex, OpenClaw, and Google Antigravity in the same license.
Also in Specs & Governance
Bundle price: $55. Compare this skill with the full workflow bundle or Pro access.
Best for
Engineers and tech leads working on non-trivial features who keep losing time to ‘the AI built the wrong thing’ — the spec is supposed to close the gap between intent and implementation, and a 10-minute structured spec routinely saves a 2-hour retrofit. Especially valuable in AI-assisted development workflows.
Not ideal for
Exploratory or research work where the right move is to write throwaway code and learn what the problem actually is — premature spec-writing locks in the wrong shape. Also a poor fit for bug fixes where the spec is implicit (‘make the test pass’) and the ceremony exceeds the work.
Included in this purchase
- Claude Code, Codex, OpenClaw, and Google Antigravity skill files.
- Setup guidance for the right adapter in your workspace.
- One-time license for the purchased skill version.
Setup
Plan for a short setup in the repository or workspace where the skill will run. Some coding familiarity helps for implementation-heavy outputs.
Related Skills
$19.99
One-time license
$19.99
One-time license
$19.99
One-time license
Future Updates
This purchase includes the current version of the skill. If you want future adapter updates — meaning compatibility and packaging updates as supported platforms evolve — plus new catalog additions included automatically, upgrade to Pro.