Component Changelog
Record what changed in the shared components so consumers are not surprised by a release.
What it adds
A dated, per-component record of additions, visual changes, deprecations, and breaking interface changes, with migration notes.
What your agent is told to do
5
What your agent is told to do
5-
1
Keep one changelog for the shared component layer and identify every entry by the component it affects, so a consumer can scan for the three components they use rather than reading the whole release.
-
2
Classify each entry by what it costs the consumer: a new component, a visual change that needs a look, a behavioural change that needs a test, a deprecation with a replacement, or a breaking interface change that needs work.
-
3
Write every deprecation with its replacement, the reason, and the release in which the old form stops working. A deprecation with no removal date is ignored indefinitely.
-
4
Generate the entries from work as it merges rather than assembling them before a release. A changelog written from memory at the end of a cycle omits precisely the small visual changes that break someone's layout.
-
5
Do not let the changelog become the specification. What a component must do lives in Component Acceptance Criteria; this record says only what changed and what the consumer has to do about it.
Edge cases it handles
7
Edge cases it handles
7- Visual and behavioural changes need separating, because they are read by different people for different reasons — a designer scans for the first, an engineer for the second, and merging them means both skim.
- A breaking change is not documented until it links to migration guidance concrete enough to follow, showing the old shape, the new shape, and what to do with anything in between.
- Theme and accessibility impact must be called out explicitly: a token that changed value, a contrast ratio that moved, an altered focus order, or a changed accessible name will not be noticed from a diff of the component.
- Entries assembled by hand at release time go stale or go missing, so derive them from merged work and treat a merged change with no entry as an incomplete change.
- A change that only affects one theme or one breakpoint must say so, or every consumer assumes it affects them and re-checks work that was never at risk.
- Internal refactors with no consumer-visible effect should be left out entirely; a changelog padded with them stops being read.
- An entry has to be tied to a version or a date that a consumer can compare against the version they are running, otherwise it cannot be acted on.
Definition of done
9
Definition of done
9- Every consumer-visible change to a shared component appears in the changelog against that component's name.
- Entries are classified as new, visual, behavioural, deprecated, or breaking.
- Each breaking change links to migration guidance showing the old and new shape.
- Each deprecation names its replacement and the release in which the old form is removed.
- Theme and accessibility impact are stated explicitly where they exist.
- Entries are produced from merged work, and a merged change without one is treated as unfinished.
- Every entry carries a version or date a consumer can compare against their own.
- The feature matches the existing design system.
- No existing functionality is broken.
Related features
AI Cost Budgets
AI Cost Budgets
Cap what AI features are allowed to spend before the bill arrives.
What it does
Monetary spending limits on AI work, scoped by workspace, feature, and time period, enforced before a run starts.
How it works
- 1 Find every place the app calls a model and route all of them through one accounting point that records estimated and actual spend against a scope. A budget that only covers the chat feature is not a budget.
- 2 Estimate the cost of a run from the size of its input before dispatching it, and refuse anything that would exceed the remaining budget on its own.
- 3 Reserve the estimate against the budget when the run starts, then reconcile to the real usage figures when it finishes, releasing whatever was over-reserved.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/ai-cost-budgets
Multi-Model Routing
Multi-Model Routing
Send each AI request to the right model using rules you can read and test.
What it does
A deterministic routing layer that picks a model per request from task type, context size, latency budget, and data sensitivity.
How it works
- 1 Express routing as explicit, ordered rules over inputs the app can measure: task type, estimated context size, latency budget, and the sensitivity classification of the data involved. A rule set that can be read line by line can be reviewed and tested.
- 2 Make routing deterministic. The same inputs must always produce the same route, so a bad output can be reproduced and a rule change can be evaluated. Randomised or load-based selection turns every incident into guesswork.
- 3 Classify data before routing and refuse to route restricted content to any destination not approved for it. This check is a hard block, not a preference, and it must run before the request is assembled.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/multi-model-routing
Model Selection
Model Selection
Let each AI task run on the model that suits its quality, speed, and cost needs.
What it does
A per-task model choice, drawn from the models the app already has configured, with capability filtering and safe defaults.
How it works
- 1 Enumerate the models the app already has access to and record what each one can actually do: context capacity, whether it can return the structured output the task requires, whether it supports the tools the task calls, and its relative cost and speed.
- 2 Offer only the models that satisfy the task's requirements. A task that needs structured output must not list a model that cannot reliably produce it, because the failure appears later as malformed responses rather than as an unavailable option.
- 3 Store the choice against the specific task, not as one global setting. A single default forces a summarisation task and a classification task onto the same tier when they have opposite needs.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/model-selection
How it works
-
1
Copy the link
Grab the Markdown instruction URL for this feature.
-
2
Give it to your AI
Paste it into Claude Code, Cursor, v0, Lovable — whatever you build with.
-
3
It inspects, then implements
Your agent reads your existing app first, then adds the feature to fit it.
Works with your stack
These instructions are written to adapt. They tell the agent to detect your framework, match your existing design system, and reuse what you already have — rather than assuming a particular stack.
Need it tighter than that? Customize the feature and tell it exactly what you're running.