feat: Add input field summaries to Actor search results

[MQ37]

Summary

• Include input field names, types, and required status in search-actors results so LLMs can prepare Actor calls without a separate fetch-actor-details round-trip
• Fetch Actor build definitions in parallel (concurrency=10) after search, with dual-layer caching (inputFieldsCache + actorDefinitionPrunedCache)
• Display as TypeScript-like notation in text: url: string, maxResults?: number
• Only show "main" input fields using sectionCaption heuristic — fields before the first section caption are the core inputs, everything after is advanced config. Reduces 20-35 fields down to 4-8 per actor.
• Show total field count when truncated: (4 of 35) so the LLM knows to use fetch-actor-details for the full schema

Text output example

- **Input fields (4 of 35):** searchStringsArray?: array, locationQuery?: string, maxCrawledPlacesPerSearch?: integer, language?: string

Structured output example

"inputFields": [
  { "name": "searchStringsArray", "type": "array", "required": false },
  { "name": "locationQuery", "type": "string", "required": false }
],
"totalInputFields": 35

Input field selection heuristic

Apify Actor input schemas use sectionCaption to group fields into UI sections. Fields before the first sectionCaption are the "main" inputs (search query, URL, limit). Fields after are advanced config (proxy, custom code, deprecated). We return only the first section. Fallback: cap at 10 fields if no sections exist.

Testing

• Type check, lint, and 84 unit tests pass
• Manually verified via mcpc against local server for Google Maps and Instagram actors

[jirispilka]

I'm afraid this PR adds a lot of complexity. It was supposed to be a simple addition to the existing logic.

The branch adds: a new cache, a generic semaphore utility, a parallel fetch pipeline that duplicates defaultBuild()+schema, and a third argument.

I can see three approaches:

1. Provide minimal code and reuse existing things as much as possible. Do not fetch inputs for all search results but for top 10. Use Promise.allSettled. Remove the semaphore. Extract helper for the defaultBuild()+schema and reuse it at the two places.

2. Add this to apify-core /store/search endpoint with responseFormat=agent. It already uses an $in join in fetchAgentSafetyContext (Acts2 + UsersActs2 + UserBackgroundChecks in parallel). Adding one more $in against ActorBuilds could be fine imo.

3. Store a pre-computed input info into the Algolia index. We would not need to hit Mongo at all as the schema would be stored there. I'm not sure how often we sync the data. So there is a risk we might get a stale input fields.

My order of preference would be 3 -> 2 -> 1 but let me pull @Jkuzz for his opinion here.

[jirispilka]

See my comment here: #737

[MQ37]

Makes sense, it is large and inefficient change as I did not want to touch the API or MongoDB - let's touch the API then.

I would be for the option 2 so we do the input fields lookup from MongoDB for the default build - no need to handle the Algolia. Right? @Jkuzz