Fix: TypeError - queries.map is not a function

[bookkeepingofficeservices-cmyk]

Fix: TypeError - queries.map is not a function

Problem

When using Perplexica with local LLMs (especially Ollama), users encounter the following error during search operations:

TypeError: a.queries.map is not a function
    at Object.execute (.next/server/chunks/641.js:41:1486)

This error occurs when the AI model returns an invalid format for the queries parameter, causing the application to crash when trying to call .slice() or .map() on a non-array value.

Root Cause

The search action handlers (webSearch, socialSearch, academicSearch) assume that input.queries is always an array, but some AI models may return:

• undefined
• null
• A string instead of an array
• An object with unexpected structure

Solution

Added defensive validation checks before processing queries:

// Fix: Ensure queries is always an array
if (!input.queries || !Array.isArray(input.queries)) {
  console.error('Invalid queries format:', input);
  input.queries = [];
}

This ensures:

1. The application doesn't crash on invalid input
2. Errors are logged for debugging
3. Search continues with an empty array (graceful degradation)

Files Changed

• src/lib/agents/search/researcher/actions/webSearch.ts
• src/lib/agents/search/researcher/actions/socialSearch.ts
• src/lib/agents/search/researcher/actions/academicSearch.ts

Testing

Tested with:

• Ollama (llama3.2:1b)
• SearXNG backend
• Various search queries in different languages

Impact

• Prevents application crashes
• Improves error handling and debugging
• Maintains backward compatibility
• No breaking changes

Related Issues

This fix addresses the common error reported by users when using local LLMs with Perplexica.

---

Summary by cubic

Prevents "TypeError: queries.map is not a function" by validating queries input and defaulting to an empty array so searches don’t crash. Sanitizes error logs and removes an accidentally committed Dockerfile.fixed.

• Bug Fixes

  • Ensure input.queries is an array before use; default to []; log only safe metadata (hasQueries, type, isArray).
  • Applied to webSearch, socialSearch, and academicSearch actions.

• Refactors

  • Remove temporary Dockerfile.fixed from the repo.

^{Written for commit cd38736. Summary will update on new commits.}

[cubic-dev-ai]

1 issue found across 3 files

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="src/lib/agents/search/researcher/actions/academicSearch.ts">

<violation number="1" location="src/lib/agents/search/researcher/actions/academicSearch.ts:35">
P2: Full payload logging in invalid-input path can leak sensitive user/model data to server logs.</violation>
</file>

---

Since this is your first cubic review, here's how it works:

• cubic automatically reviews your code and comments on bugs and improvements
• Teach cubic by replying to its comments. cubic learns from your replies and gets better over time
• Add one-off context when rerunning by tagging @cubic-dev-ai with guidance or docs links (including llms.txt)
• Ask questions if you need clarification on any suggestion

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[bookkeepingofficeservices-cmyk]

Thanks for the security review! I've updated the code to avoid logging sensitive data. Now only logging safe metadata (query type, array status) for debugging purposes.

Changes made:

• Replaced console.error('Invalid queries format:', input) with structured logging
• Only log: hasQueries, queriesType, and isArray flags
• No sensitive user or model data is exposed

Commit: cd5cc3c

[cubic-dev-ai]

2 issues found across 4 files (changes from recent commits).

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="Dockerfile.fixed">

<violation number="1" location="Dockerfile.fixed:1">
P2: Docker base image uses mutable `latest` tag, making builds non-reproducible and susceptible to unreviewed upstream changes.</violation>

<violation number="2" location="Dockerfile.fixed:9">
P2: The rebuild guard is ineffective: checking only `/home/perplexica-src` directory existence is always true after the preceding COPY commands, so `npm run build` runs unconditionally.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}