ClawTrust
1.7k Downloads
Overview
Help the agent design, implement, and maintain Raycast extensions using React, TypeScript, and the Raycast API, guided by the repo’s references/api/*.md specification files.
Key Advantages
1.Tightly focused on Raycast extension development with explicit support for List, Grid, Detail, and Form components.
2.Encodes a clear workflow (identify component, consult reference, implement with defaults) that reduces common Raycast UI/API mistakes.
3.Uses the repo’s references/api/*.md as the primary source of truth, improving alignment with documented Raycast APIs.
4.Promotes best practices for feedback (showToast vs showHUD) and data persistence (Cache vs LocalStorage).
5.Includes patterns for restricted APIs (AI, BrowserExtension) with environment.canAccess checks to avoid invalid calls.
Use Cases
- Creating new Raycast commands that display data in List or Grid with search, actions, and navigation.
- Building rich detail views using Detail with markdown and metadata for item inspection or documentation-style displays.
- Implementing interactive forms (Form) that collect user input and submit it via Action.SubmitForm handlers.
- Adding user feedback flows using showToast for loading, success, and failure states, and showHUD for quick background notifications.
- Managing transient vs persistent data via Cache and LocalStorage for performance and user preference storage in Raycast extensions. Integrating AI.ask and BrowserExtension APIs in commands while obeys
Evaluation Scores
8.3
/ 10
Reliability
7.5
Functionality
8.0
Usability
8.5
Safety
9.0
Performance
8.5
Compatibility
8.5
Based on 1 evaluation · Latest: 3/19/2026
Download Trend
Loading...
Evaluation History (1)
8.3/103/19/2026▼
OS: win32-x64LLM: arcee-ai/trinity-large-preview
**Quick judgment**
Specialized and likely very effective for Raycast extension work. It encodes Raycast-specific patterns, references the repo’s `references/api/*.md` files as ground truth, and guides the agent toward idiomatic use of List, Grid, Detail, Form, feedback APIs, caching, and restricted APIs. Best suited when the task is clearly scoped to Raycast extension UI or behavior in a project containing the documented reference files.
**Key strengths**
- Focus on core Raycast UI components (List, Grid, Detail, Form) with cookbook-style patterns.
- Encourages correct feedback mechanisms (showToast vs showHUD) and data persistence choices (Cache vs LocalStorage).
- Enforces environment checks before using AI and BrowserExtension, reducing invalid or unsafe API usage.
- Uses local `references/api/*.md` specs as the main authority, which should keep implementations aligned with documented APIs when those files are up to date.
**Risks and limitations**
- Highly domain-specific: not useful outside Raycast extension development and assumes the presence and correctness of the local `references/api/*.md` files.
- If the reference files are outdated relative to the actual Raycast API, generated code could be subtly incorrect (types, props, or behaviors).
- Cookbook patterns are helpful but may not cover advanced or edge-case behaviors; complex extensions may still need manual refinement.
- Safety risks are low overall, but misuse of AI or BrowserExtension APIs is still possible if environment checks are omitted or the surrounding project logic is insecure.
**Recommended scenarios**
- Implementing or fixing UI and logic for Raycast commands that use List, Grid, Detail, or Form.
- Adding toasts/HUD feedback and simple persistence to existing Raycast extensions.
- Rapidly prototyping Raycast extensions when you have the repo (with `references/api/*.md`) locally available and want idiomatic, spec-aligned code.
- Teaching or standardizing Raycast patterns within a team, using the skill’s workflow (identify component → consult reference → implement with defaults).
Comments (0)
No comments yet. Be the first!