Rename "planetary copilot" to "planet computer" and clear example prompts#645
Rename "planetary copilot" to "planet computer" and clear example prompts#645ngoiyaeric wants to merge 5 commits into
Conversation
…mpts - Replaced "planetary copilot" with "planet computer" in all settings views and forms. - Updated system prompt placeholder to "Enter your planet computer system prompt". - Set default system prompt to an empty string. - Adjusted Zod validation for system prompt to allow empty strings. - Emptied the `exampleMessages` array in `EmptyScreen`. Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
WalkthroughThis PR updates terminology from "planetary copilot" to "planet computer" across the settings UI, allows empty system prompts via relaxed validation, adds map-provider-aware gating to search button availability, and implements vision-capability-aware model selection for xAI and Gemini providers. ChangesSettings, Search & Model Selection Updates
Estimated code review effort🎯 2 (Simple) | ⏱️ ~12 minutes Possibly related issues
Possibly related PRs
Suggested labels
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Warning There were issues while running some tools. Please review the errors and either fix the tool's configuration or disable the tool if it's a critical failure. 🔧 ESLint
ESLint install failed. For unrecoverable errors, disable the tool in CodeRabbit configuration. 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 |
Review Summary by QodoRename planetary copilot to planet computer and clear defaults
WalkthroughsDescription• Rename "planetary copilot" to "planet computer" across all settings UI • Clear default system prompt and example messages from empty screen • Allow empty system prompt by updating validation to min(0) • Update all related placeholders and descriptions for consistency Diagramflowchart LR
A["Settings Components"] -->|"Rename terminology"| B["planet computer"]
C["System Prompt"] -->|"Clear default"| D["Empty string"]
E["Validation Schema"] -->|"Allow empty"| F["min(0)"]
G["Empty Screen"] -->|"Clear examples"| H["Empty array"]
File Changes1. components/empty-screen.tsx
|
Code Review by QodoLooking for bugs?Check back in a few minutes. An AI review agent is analyzing this pull request.ⓘ You are approaching your monthly quota for Qodo. Upgrade your plan |
… default prompt - Replaced "planetary copilot" with "planet computer" in settings views and forms. - Set default system prompt to empty string. - Updated system prompt placeholder to "Enter your planet computer system prompt". - Kept example messages in EmptyScreen (reverted previous accidental removal). Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@components/settings/components/settings.tsx`:
- Around line 28-34: The frontend allows an empty systemPrompt but
saveSystemPrompt (in lib/actions/chat.ts) rejects falsy prompts; update
saveSystemPrompt to accept empty string values by either treating an empty
string as valid or converting empty strings to null before validation/storage so
clearing the prompt succeeds; specifically, modify the saveSystemPrompt
function's input check (the if (!prompt) early return) to allow "" (or normalize
"" -> null) and proceed to persist that normalized value, ensuring callers like
app/actions.tsx and lib/agents/researcher.tsx still get the expected fallback
behavior.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: ASSERTIVE
Plan: Pro
Run ID: 3fd5c121-3a57-44a0-a836-9daf3e9f28da
📒 Files selected for processing (5)
components/empty-screen.tsxcomponents/settings/components/model-selection-form.tsxcomponents/settings/components/settings.tsxcomponents/settings/components/system-prompt-form.tsxcomponents/settings/settings-view.tsx
📜 Review details
🧰 Additional context used
🧠 Learnings (1)
📚 Learning: 2026-04-25T08:10:37.673Z
Learnt from: ngoiyaeric
Repo: QueueLab/QCX PR: 0
File: :0-0
Timestamp: 2026-04-25T08:10:37.673Z
Learning: In the QCX repository (feat/ai-sdk-ui-migration branch and going forward), there is no token generation/counting tracking and no multimodal LLM features. Base64 image data sent via `append()` body fields in components like `header-search-button.tsx` and `resolution-carousel.tsx` is for server-side `resolution_search` processing only — it is NOT passed as multimodal message content parts to a vision model. Avoid flagging token tracking or multimodal LLM handling gaps as review concerns in this codebase.
Applied to files:
components/empty-screen.tsx
🔇 Additional comments (6)
components/empty-screen.tsx (1)
4-4: LGTM!components/settings/components/settings.tsx (2)
52-56: LGTM!
176-186: LGTM!Also applies to: 188-198
components/settings/components/system-prompt-form.tsx (1)
13-35: LGTM!components/settings/components/model-selection-form.tsx (1)
136-140: LGTM!components/settings/settings-view.tsx (1)
18-29: LGTM!
| const settingsFormSchema = z.object({ | ||
| systemPrompt: z | ||
| .string() | ||
| .min(10, { | ||
| message: "System prompt must be at least 10 characters.", | ||
| }) | ||
| .min(0) | ||
| .max(2000, { | ||
| message: "System prompt cannot exceed 2000 characters.", | ||
| }), |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Description: Verify handling of empty system prompts in saveSystemPrompt and chat logic
# Search for saveSystemPrompt implementation
echo "=== saveSystemPrompt implementation ==="
ast-grep --pattern 'export async function saveSystemPrompt($$$) {
$$$
}'
# Search for system prompt usage in chat/AI logic
echo -e "\n=== System prompt usage in chat logic ==="
rg -n -A 5 -B 2 'systemPrompt' --type ts --glob '**/chat/**' --glob '**/ai/**' --glob '**/lib/actions/**'
# Look for potential string concatenation or template usage
echo -e "\n=== Potential template/concatenation usage ==="
rg -n -P 'systemPrompt.*\+|`.*\$\{.*systemPrompt.*\}' --type tsRepository: QueueLab/QCX
Length of output: 2454
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== Locate saveSystemPrompt / getSystemPrompt definitions ==="
rg -n --hidden --no-ignore -S "function saveSystemPrompt|saveSystemPrompt\s*\(|function getSystemPrompt|getSystemPrompt\s*\(" -t ts -g '!*node_modules*'
echo -e "\n=== Inspect lib/actions/chat.ts around getSystemPrompt/saveSystemPrompt and model usage ==="
FILE="lib/actions/chat.ts"
if [ -f "$FILE" ]; then
nl -ba "$FILE" | sed -n '1,260p'
else
echo "Missing $FILE"
fi
echo -e "\n=== Inspect settings schema and defaults in components/settings/components/settings.tsx ==="
SFILE="components/settings/components/settings.tsx"
if [ -f "$SFILE" ]; then
nl -ba "$SFILE" | sed -n '1,160p'
else
echo "Missing $SFILE"
fi
echo -e "\n=== Other systemPrompt usages in AI/chat code paths ==="
rg -n -S "systemPrompt" -t ts -g 'lib/**' -g 'components/**' -g 'app/**' | head -n 200Repository: QueueLab/QCX
Length of output: 669
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== lib/actions/chat.ts: saveSystemPrompt & getSystemPrompt ==="
sed -n '130,210p' lib/actions/chat.ts
echo -e "\n=== app/actions.tsx: currentSystemPrompt handling ==="
sed -n '430,540p' app/actions.tsx
echo -e "\n=== components/settings/components/settings.tsx: schema + default ==="
sed -n '1,160p' components/settings/components/settings.tsx
echo -e "\n=== Find where systemPrompt/currentSystemPrompt is passed to the model ==="
rg -n -S "currentSystemPrompt|systemPrompt\s*[:)]|system:\s*.*(systemPrompt|currentSystemPrompt)" -t ts app lib components
echo -e "\n=== Find any normalization/fallback for empty systemPrompt ==="
rg -n -S "systemPrompt.*\|\|\s*''|systemPrompt.*??|if\s*\(.*systemPrompt|trim\(\).*systemPrompt" -t ts app lib componentsRepository: QueueLab/QCX
Length of output: 12654
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "=== lib/agents/researcher.tsx: how systemPromptToUse is derived ==="
sed -n '1,220p' lib/agents/researcher.tsx
echo -e "\n=== lib/agents/resolution-search.tsx: system prompt handling ==="
sed -n '1,220p' lib/agents/resolution-search.tsx
echo -e "\n=== components/settings/components/system-prompt-form.tsx (UI constraints) ==="
sed -n '1,120p' components/settings/components/system-prompt-form.tsxRepository: QueueLab/QCX
Length of output: 14997
🏁 Script executed:
#!/bin/bash
set -euo pipefail
sed -n '1,120p' lib/db/schema.ts
sed -n '120,200p' lib/db/schema.tsRepository: QueueLab/QCX
Length of output: 6945
Fix empty system prompt persistence mismatch
The form schema allows an empty systemPrompt (.min(0) with default ""), but lib/actions/chat.ts’s saveSystemPrompt rejects it via if (!prompt) return { error: 'Prompt is required' }, so clearing/saving an empty prompt fails.
Chat/AI runtime is likely safe because app/actions.tsx converts null to '', and lib/agents/researcher.tsx falls back to getDefaultSystemPrompt(...) when dynamicSystemPrompt.trim() is empty.
Update saveSystemPrompt to accept empty strings (or convert empty string to null for storage) so UI behavior matches backend behavior.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@components/settings/components/settings.tsx` around lines 28 - 34, The
frontend allows an empty systemPrompt but saveSystemPrompt (in
lib/actions/chat.ts) rejects falsy prompts; update saveSystemPrompt to accept
empty string values by either treating an empty string as valid or converting
empty strings to null before validation/storage so clearing the prompt succeeds;
specifically, modify the saveSystemPrompt function's input check (the if
(!prompt) early return) to allow "" (or normalize "" -> null) and proceed to
persist that normalized value, ensuring callers like app/actions.tsx and
lib/agents/researcher.tsx still get the expected fallback behavior.
|
@coderabbitai the resolution search is not working but there are no code changes regarding that, can you debug that |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
# Search for resolution search related code
echo "=== Searching for 'resolution' in source files ==="
rg -rn "resolution" --type ts --type tsx -C 3 2>/dev/null | head -200
echo ""
echo "=== Searching for 'resolution' in all files ==="
rg -rn "resolution" -C 3 --glob "!node_modules" --glob "!.git" --glob "!*.lock" 2>/dev/null | head -300
echo ""
echo "=== Searching for 'search' related API routes ==="
fd -e ts -e tsx --exclude node_modules | xargs grep -l "search" 2>/dev/null | head -30Length of output: 12754 🏁 Script executed: #!/bin/bash
echo "=== File structure overview ==="
tree -L 3 --gitignore 2>/dev/null | head -100
echo ""
echo "=== API routes ==="
fd -t f -e ts --exclude node_modules -p "app/api" 2>/dev/null
find . -path ./node_modules -prune -o -path "./.git" -prune -o -name "*.ts" -path "*/api/*" -print 2>/dev/null | head -50Length of output: 3149 🏁 Script executed: #!/bin/bash
echo "=== Search for 'resolution' keyword broadly ==="
rg -rn "resolution" --glob "!node_modules" --glob "!.git" --glob "!*.lock" --glob "!*.json" -l 2>/dev/null
echo ""
echo "=== Search for search-related components or hooks ==="
rg -rn "resolution" --glob "!node_modules" --glob "!.git" --glob "!*.lock" -C 5 2>/dev/null | head -400Length of output: 17720 🏁 Script executed: #!/bin/bash
echo "=== Check if n-search.tsx exists vs resolution-search.tsx ==="
fd -t f "n-search" --exclude node_modules 2>/dev/null
fd -t f "resolution-search" --exclude node_modules 2>/dev/null
echo ""
echo "=== Check if n-carousel exists ==="
fd -t f "n-carousel" --exclude node_modules 2>/dev/null
fd -t f "n-image" --exclude node_modules 2>/dev/null
fd -t f "resolution-carousel" --exclude node_modules 2>/dev/null
fd -t f "resolution-image" --exclude node_modules 2>/dev/null
echo ""
echo "=== Check n-search schema vs resolution-search schema ==="
fd -t f "n-search" lib/schema --exclude node_modules 2>/dev/null
fd -t f -e ts -e tsx lib/schema --exclude node_modules 2>/dev/null
echo ""
echo "=== Full content of lib/agents/resolution-search.tsx ==="
cat lib/agents/resolution-search.tsx
echo ""
echo "=== Full content of lib/agents/index.tsx ==="
cat lib/agents/index.tsx
echo ""
echo "=== Full content of lib/schema/resolution-search.ts ==="
cat lib/schema/resolution-search.ts
echo ""
echo "=== Imports in app/actions.tsx ==="
head -40 app/actions.tsxLength of output: 11964 🏁 Script executed: #!/bin/bash
echo "=== Full resolution search invocation in app/actions.tsx ==="
grep -n "nSearch\|resolutionSearch\|n_search\|processResolution" app/actions.tsx
echo ""
echo "=== Full content of components/header-search-button.tsx ==="
cat components/header-search-button.tsx
echo ""
echo "=== Full content of components/resolution-carousel.tsx ==="
cat components/resolution-carousel.tsx
echo ""
echo "=== Check what's actually exported from lib/agents/resolution-search.tsx ==="
grep -n "^export" lib/agents/resolution-search.tsx
echo ""
echo "=== Check what nSearch resolves to in lib/agents/index.tsx ==="
cat lib/agents/index.tsx
echo ""
echo "=== Does n-search file exist anywhere? ==="
fd "n-search" --exclude node_modules 2>/dev/null
echo ""
echo "=== Actions.tsx: full n_search action block ==="
grep -n -A 80 "if (action === 'n_search')" app/actions.tsx | head -120Length of output: 14788
Root CauseThe search button's // Line ~163 (desktop button) and ~175 (mobile button)
// BUG: !map is always true for Google Maps users
disabled={isAnalyzing || !map || !actions}The handler function already handles this correctly with a provider guard: if (mapProvider === 'mapbox' && !map) {
toast.error('Map is not available yet...')
return
}But the Fix// Desktop button
disabled={isAnalyzing || (mapProvider === 'mapbox' && !map) || !actions}
// Mobile button
disabled={isAnalyzing || (mapProvider === 'mapbox' && !map) || !actions}This ensures:
The rest of the resolution search pipeline — imports, action string ( Would you like me to open a GitHub issue or plan the code fix for this? |
I've investigated the issue with Resolution Search. It was failing because the |
…ter" - Updated `getModel` in `lib/utils/index.ts` to support vision models (`grok-vision-beta`, `gemini-1.5-pro`) when `requireVision` is true. - Replaced "planetary copilot" with "planet computer" in all settings views and forms. - Updated system prompt placeholder to "Enter your planet computer system prompt". - Set default system prompt to an empty string. - Adjusted Zod validation for system prompt to allow empty strings. - Restored example messages in `EmptyScreen`. Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
- Fixed `HeaderSearchButton` being disabled for Google Maps users by updating the `disabled` prop logic. - Updated `getModel` to correctly select vision models (`grok-vision-beta`, `gemini-1.5-pro`) when `requireVision` is true. - Renamed "planetary copilot" to "planet computer" in settings. - Cleared default system prompt and updated placeholder. - Updated Zod validation for system prompt. - Restored example messages in `EmptyScreen`. Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
|
@coderabbitai the resolution search is not working but there are no code changes regarding that, can you debug that |
|
Tip For best results, initiate chat on the files or code changes.
Root Cause —
|
I've identified and fixed two issues that were preventing Resolution Search from working correctly:
These changes are included in this PR update. |
- Fix getModel to return vision-capable models when requireVision is true. - Fix HeaderSearchButton disabled state for Google Maps users. - Rename "planetary copilot" to "planet computer" in settings. - Empty default system prompt and update placeholder. - Restore example messages in EmptyScreen. Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@lib/utils/index.ts`:
- Line 45: getModel currently only switches on requireVision and returns
xai("grok-4-fast-non-reasoning") for non-vision use, which is unsafe for
structured-output; add a new boolean parameter (e.g., requireStructuredOutput or
supportsJsonSchema) to the getModel signature in lib/utils/index.ts and when
true return a JSON-schema-capable model (e.g., openai("gpt-4o") or another
approved model) instead of grok-4-fast-non-reasoning; update all call sites that
use generateObject/streamObject with a schema (e.g., generateObject({ model:
(await getModel()), schema: nextActionSchema }) in lib/agents/task-manager.tsx
and streamObject calls in lib/agents/query-suggestor.tsx,
lib/agents/inquire.tsx, lib/agents/resolution-search.tsx and
lib/actions/suggest.ts) to pass the new flag so structured-output flows select
the JSON-schema-capable model while leaving vision and other paths unchanged.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: ASSERTIVE
Plan: Pro
Run ID: 9752be9f-a43b-4583-935b-a2c4729e552d
📒 Files selected for processing (2)
components/header-search-button.tsxlib/utils/index.ts
📜 Review details
🧰 Additional context used
🧠 Learnings (3)
📚 Learning: 2026-04-25T08:10:37.673Z
Learnt from: ngoiyaeric
Repo: QueueLab/QCX PR: 0
File: :0-0
Timestamp: 2026-04-25T08:10:37.673Z
Learning: In the QCX repository (feat/ai-sdk-ui-migration branch and going forward), there is no token generation/counting tracking and no multimodal LLM features. Base64 image data sent via `append()` body fields in components like `header-search-button.tsx` and `resolution-carousel.tsx` is for server-side `resolution_search` processing only — it is NOT passed as multimodal message content parts to a vision model. Avoid flagging token tracking or multimodal LLM handling gaps as review concerns in this codebase.
Applied to files:
components/header-search-button.tsx
📚 Learning: 2026-01-13T13:26:30.086Z
Learnt from: ngoiyaeric
Repo: QueueLab/QCX PR: 0
File: :0-0
Timestamp: 2026-01-13T13:26:30.086Z
Learning: Workers that rely on ai.generateObject (geojsonParser, map-command-generator, feedback-analyzer) must not use the X.AI 'grok-4-fast-non-reasoning' model because X.AI’s chat/completions rejects JSON Schema structured-output arguments; prefer OpenAI gpt-4o for structured outputs.
Applied to files:
lib/utils/index.ts
📚 Learning: 2026-01-17T06:14:51.070Z
Learnt from: ngoiyaeric
Repo: QueueLab/QCX PR: 0
File: :0-0
Timestamp: 2026-01-17T06:14:51.070Z
Learning: For structured output tasks using ai.generateObject (resolution-search, geojsonParser, map-command-generator, feedback-analyzer), prefer OpenAI gpt-4o. While xAI's grok-4-1-fast-reasoning technically supports structured outputs and vision, OpenAI has proven more reliable integration with the AI SDK's generateObject function and Zod schemas in production.
Applied to files:
lib/utils/index.ts
🔇 Additional comments (1)
components/header-search-button.tsx (1)
174-174: LGTM!Also applies to: 186-186
| }); | ||
| try { | ||
| return xai('grok-4-fast-non-reasoning'); | ||
| return xai(requireVision ? "grok-vision-beta" : "grok-4-fast-non-reasoning"); |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Verify where getModel() is used with structured-output APIs.
# Expected: identify call sites where getModel(...) feeds streamObject/generateObject flows.
set -euo pipefail
echo "== getModel call sites =="
rg -n -C2 '\bgetModel\s*\(' --type ts --type tsx
echo
echo "== structured output API usage (streamObject/generateObject) =="
rg -n -C3 '\b(streamObject|generateObject)\s*\(' --type ts --type tsx
echo
echo "== resolution-search and related agents around model selection =="
rg -n -C4 'model:\s*await\s+getModel|model:\s*\(await getModel' lib --type ts --type tsxRepository: QueueLab/QCX
Length of output: 128
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "== getModel call sites =="
rg -n -C2 '\bgetModel\s*\(' lib -g'*.ts' -g'*.tsx'
echo
echo "== structured output call sites =="
rg -n -C3 '\b(streamObject|generateObject)\s*\(' lib -g'*.ts' -g'*.tsx'
echo
echo "== look for model: (await) getModel near structured output usage =="
rg -n -C4 'model:\s*\(?\s*(await\s+)?getModel' lib -g'*.ts' -g'*.tsx'Repository: QueueLab/QCX
Length of output: 10620
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "== lib/utils/index.ts around getModel =="
sed -n '1,140p' lib/utils/index.ts | nl -ba | sed -n '1,140p'
echo
echo "== lib/utils/index.ts: lines ~35-110 (for exact grok routing) =="
sed -n '35,110p' lib/utils/index.ts | nl -ba
echo
echo "== lib/agents/task-manager.tsx: generateObject block =="
sed -n '1,120p' lib/agents/task-manager.tsx | nl -ba | sed -n '1,120p'
echo
echo "== lib/agents/query-suggestor.tsx: streamObject block end (schema) =="
sed -n '1,120p' lib/agents/query-suggestor.tsx | nl -ba | sed -n '40,120p'
echo
echo "== lib/agents/inquire.tsx: streamObject block end (schema) =="
sed -n '1,120p' lib/agents/inquire.tsx | nl -ba | sed -n '1,120p'Repository: QueueLab/QCX
Length of output: 158
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "== lib/utils/index.ts (first 140 lines) =="
sed -n '1,140p' lib/utils/index.ts | cat -n
echo
echo "== lib/agents/task-manager.tsx (first 140 lines) =="
sed -n '1,140p' lib/agents/task-manager.tsx | cat -n
echo
echo "== lib/agents/query-suggestor.tsx (first 120 lines) =="
sed -n '1,120p' lib/agents/query-suggestor.tsx | cat -n
echo
echo "== lib/agents/inquire.tsx (first 120 lines) =="
sed -n '1,120p' lib/agents/inquire.tsx | cat -nRepository: QueueLab/QCX
Length of output: 15644
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "== lib/utils/index.ts: grok-4-fast-non-reasoning routing =="
rg -n "grok-4-fast-non-reasoning" lib/utils/index.ts
echo
echo "== structured output call sites using getModel() =="
rg -n -C2 "model:\s*\(await\s+getModel\(\)\)" lib -g'*.ts' -g'*.tsx' | rg -n -C2 "(streamObject|generateObject)"
echo
echo "== structured output call sites (streamObject/generateObject) that pass schema =="
rg -n -C2 "(streamObject|generateObject)\s*\(" lib -g'*.ts' -g'*.tsx' | rg -n -C2 "schema:"Repository: QueueLab/QCX
Length of output: 2028
Route structured-output calls to a JSON-schema-compatible model (xAI grok-4-fast-non-reasoning is unsafe)
lib/utils/index.ts selects xAI grok-4-fast-non-reasoning whenever requireVision is false (lines 45 and 90), but multiple structured-output call sites use generateObject/streamObject with a schema while calling getModel() with the default requireVision=false, e.g.:
lib/agents/task-manager.tsx:generateObject({ model: (await getModel()), schema: nextActionSchema })lib/agents/query-suggestor.tsx:streamObject({ model: (await getModel()), schema: relatedSchema })lib/agents/inquire.tsx:streamObject({ model: (await getModel()), schema: inquirySchema })- (Also)
lib/agents/resolution-search.tsx/lib/actions/suggest.ts:streamObject(..., schema: ...)withgetModel(hasImage)(so non-image paths can still hitgrok-4-fast-non-reasoning)
Per the guidance, structured-output workers must not use grok-4-fast-non-reasoning; model selection needs a dedicated “structured-output/JSON-schema capable” signal (not just requireVision) and must route those calls to a compatible model (e.g., OpenAI gpt-4o).
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@lib/utils/index.ts` at line 45, getModel currently only switches on
requireVision and returns xai("grok-4-fast-non-reasoning") for non-vision use,
which is unsafe for structured-output; add a new boolean parameter (e.g.,
requireStructuredOutput or supportsJsonSchema) to the getModel signature in
lib/utils/index.ts and when true return a JSON-schema-capable model (e.g.,
openai("gpt-4o") or another approved model) instead of
grok-4-fast-non-reasoning; update all call sites that use
generateObject/streamObject with a schema (e.g., generateObject({ model: (await
getModel()), schema: nextActionSchema }) in lib/agents/task-manager.tsx and
streamObject calls in lib/agents/query-suggestor.tsx, lib/agents/inquire.tsx,
lib/agents/resolution-search.tsx and lib/actions/suggest.ts) to pass the new
flag so structured-output flows select the JSON-schema-capable model while
leaving vision and other paths unchanged.
1 participant
This PR renames all occurrences of "planetary copilot" to "planet computer" within the settings section of the application. It also sets the default system prompt to an empty string and updates the placeholder to "Enter your planet computer system prompt". Furthermore, it removes all example prompt messages from the empty screen as requested.
Key changes:
settings-view.tsx,settings.tsx,system-prompt-form.tsx, andmodel-selection-form.tsx.settingsFormSchemainsettings.tsxto allow amin(0)length for the system prompt.exampleMessagesinempty-screen.tsx.PR created automatically by Jules for task 13200136455343942810 started by @ngoiyaeric
Summary by CodeRabbit
New Features
Bug Fixes
Style