fix: update prompts #2910
fix: update prompts #2910
Conversation
|
This pull request has been ignored for the connected project Preview Branches by Supabase. |
WalkthroughConsolidates imports around a new constants entry point, updates prompt text in context and system constants, adds a constants barrel export, adjusts the prompt package’s public exports, and updates dependent import paths. One web API router merges imports. No runtime control-flow changes. Changes
Sequence Diagram(s)Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Possibly related PRs
Poem
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
apps/web/client/src/server/api/routers/chat/suggestion.ts (2)
13-19: Block user-supplied system-role messages (prompt injection vector).Allowing clients to send role: "system" after your own system prompt can override it. Restrict roles to user/assistant and filter out any incoming system messages.
Apply these diffs:
- messages: z.array(z.object({ - role: z.enum(['user', 'assistant', 'system']), + messages: z.array(z.object({ + role: z.enum(['user', 'assistant']), content: z.string(), })),- ...convertToModelMessages(input.messages.map((m) => ({ + ...convertToModelMessages( + input.messages + .filter((m) => m.role !== 'system') + .map((m) => ({ role: m.role, parts: [{ type: 'text', text: m.content }], - }))), + }))),Also applies to: 34-37
47-49: Restrict suggestion update to current project
Add a predicate on conversations.projectId to match ctx.session.projectId and import and:-import { eq } from 'drizzle-orm'; +import { and, eq } from 'drizzle-orm'; … - await ctx.db.update(conversations).set({ - suggestions, - }).where(eq(conversations.id, input.conversationId)); + await ctx.db.update(conversations).set({ suggestions }).where( + and( + eq(conversations.id, input.conversationId), + eq(conversations.projectId, ctx.session.projectId) + ) + );
🧹 Nitpick comments (2)
packages/ai/src/prompt/constants/context.ts (2)
1-1: Minor copy edit (optional).Consider making the prefix more direct; current copy is fine if intentionally conversational.
24-24: Brand capitalization nit (optional).Use “Next.js” and “Tailwind CSS” for consistency.
-const projectContextPrefix = `This is a Nextjs project with TailwindCSS`; +const projectContextPrefix = `This is a Next.js project with Tailwind CSS`;
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (7)
apps/web/client/src/server/api/routers/chat/suggestion.ts(1 hunks)packages/ai/src/prompt/constants/context.ts(1 hunks)packages/ai/src/prompt/constants/index.ts(1 hunks)packages/ai/src/prompt/constants/system.ts(1 hunks)packages/ai/src/prompt/index.ts(1 hunks)packages/ai/src/prompt/provider.ts(1 hunks)packages/ai/src/tools/classes/onlook-instructions.ts(1 hunks)
🧰 Additional context used
📓 Path-based instructions (4)
**/*.{ts,tsx}
📄 CodeRabbit inference engine (AGENTS.md)
Do not use the any type unless necessary
Files:
packages/ai/src/prompt/index.tspackages/ai/src/prompt/constants/index.tsapps/web/client/src/server/api/routers/chat/suggestion.tspackages/ai/src/prompt/constants/context.tspackages/ai/src/prompt/constants/system.tspackages/ai/src/prompt/provider.tspackages/ai/src/tools/classes/onlook-instructions.ts
{apps,packages}/**/*.{ts,tsx}
📄 CodeRabbit inference engine (CLAUDE.md)
Avoid using the any type unless absolutely necessary
Files:
packages/ai/src/prompt/index.tspackages/ai/src/prompt/constants/index.tsapps/web/client/src/server/api/routers/chat/suggestion.tspackages/ai/src/prompt/constants/context.tspackages/ai/src/prompt/constants/system.tspackages/ai/src/prompt/provider.tspackages/ai/src/tools/classes/onlook-instructions.ts
apps/web/client/src/server/api/routers/**/*.ts
📄 CodeRabbit inference engine (AGENTS.md)
apps/web/client/src/server/api/routers/**/*.ts: Place tRPC routers under apps/web/client/src/server/api/routers/**
Use publicProcedure/protectedProcedure from apps/web/client/src/server/api/trpc.ts and validate inputs with Zod
Return plain objects/arrays; rely on SuperJSON for serialization in tRPC procedures
apps/web/client/src/server/api/routers/**/*.ts: Place tRPC routers under src/server/api/routers/**
Use publicProcedure/protectedProcedure from src/server/api/trpc.ts and validate inputs with Zod
Return plain objects/arrays; rely on SuperJSON for serialization
Files:
apps/web/client/src/server/api/routers/chat/suggestion.ts
apps/web/client/src/**/*.{ts,tsx}
📄 CodeRabbit inference engine (AGENTS.md)
apps/web/client/src/**/*.{ts,tsx}: Use path aliases @/* and ~/* for imports that map to apps/web/client/src/*
Avoid hardcoded user-facing text; use next-intl messages/hooks insteadUse path aliases @/* and ~/* for imports mapping to src/*
Files:
apps/web/client/src/server/api/routers/chat/suggestion.ts
🔇 Additional comments (7)
apps/web/client/src/server/api/routers/chat/suggestion.ts (2)
1-1: Import consolidation LGTMSingle-source import from @onlook/ai improves clarity and aligns with the new public surface.
43-44: Verify token limit option name and right-sizing.For OpenRouter via the ai SDK, the option is typically maxTokens, not maxOutputTokens. 10k is excessive for 3 brief suggestions.
Consider:
- maxOutputTokens: 10000, + maxTokens: 512,If your ai package expects maxOutputTokens, keep it; otherwise switch to maxTokens. Please confirm against your installed ai SDK.
packages/ai/src/tools/classes/onlook-instructions.ts (1)
4-4: Import path update LGTMSwitching ONLOOK_INSTRUCTIONS to the constants barrel matches the new public surface.
packages/ai/src/prompt/constants/system.ts (1)
14-14: Directive tightening LGTMStronger guidance around data-oid handling is clear and unambiguous.
packages/ai/src/prompt/provider.ts (1)
10-10: Barrel import consolidation LGTMReduces import churn and centralizes prompt sourcing.
packages/ai/src/prompt/constants/index.ts (1)
1-11: New constants barrel LGTMCentralized re-exports simplify consumption and align with related import changes.
packages/ai/src/prompt/index.ts (1)
1-1: No stale imports remain. Search across the codebase found no references toprompt/create,prompt/onlook, orprompt/summary. Public surface change is safe.
Description
Related Issues
Type of Change
Testing
Screenshots (if applicable)
Additional Notes
Important
Refactor prompt files into a
constantsdirectory and update related imports, with minor prompt content updates for clarity.constantsdirectory inpackages/ai/src/prompt/.provider.ts,onlook-instructions.ts, andsuggestion.tsto reflect new paths.filesContentPrefixandhighlightPrefixincontext.tsfor clarity.SYSTEM_PROMPTinsystem.tsto emphasize not alteringdata-oidattributes.index.tsinconstantsdirectory to export all constants.This description was created by
for 700c721. You can customize this summary. It will automatically update as commits are pushed.
Summary by CodeRabbit
Documentation
Refactor