Normalize cloud-function strategy handling, add StrategyKey typing, sync selection flow and UI updates

[esclipse]

Motivation

• Ensure the client-selected strategy is respected and mapped to the cloud function key instead of always requesting all strategies.
• Normalize multi-strategy cloud results into a consistent matches shape with strategy tags and deduplication so the frontend can present merged results.
• Replace the old polling proxy with a synchronous request flow that includes retries, abort timeouts, and robust response parsing.
• Fix TypeScript typing issues (duplicate strategy identifier) and preserve optional AI scoring while exposing a maxDuration constant.

Description

• Add a StrategyKey union type and make normalizeStrategy explicitly return StrategyKey, and use it in convertCloudFunctionResult.
• Replace polling/proxy logic with a synchronous POST flow that calls the cloud function, uses parseCloudFunctionResponse, retry/backoff, request timeouts/abort, and returns normalized { matches, stats, summary }.
• Normalize and deduplicate results from results.strategy1|2|3, add optional fields like strategy, j_value, and j_last on match objects, and compute stats/summary.
• Update UI in app/stock/page.tsx to remove polling, add symbolsInput and summary state, make some match fields optional in types, and render improved loading/empty/error and AI analysis output.

Testing

• The TypeScript compiler was run and initially failed due to a duplicate strategy identifier, which was corrected by introducing the StrategyKey type (compile failure observed before the fix).
• No other automated unit or integration tests were executed for these changes.
• Manual safeguards (timeouts/retries) were exercised during development but are not part of automated tests.

---

Codex Task

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Review	Updated (UTC)
mcp-test	Ready	Preview, Comment	Jan 9, 2026 9:53am

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4d563acf1f

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Align full-analysis timeout with 5‑minute route limit

When no symbols are provided, the handler sets a 15‑minute timeout (timeout = ... 15 * 60 * 1000), but the route declares maxDuration = 300 (5 minutes). In that scenario (full analysis without symbols), the serverless runtime will terminate the request before your 15‑minute abort/retry loop can finish, so the client will never receive a successful response even if the cloud function eventually completes. This regression is specific to long‑running full scans after removing polling; consider lowering the timeout to ≤300s, re‑introducing polling, or increasing maxDuration if the platform supports it.

Useful? React with 👍 / 👎.