fix(caido): normalize string null and sentinels for scopeId and parentId (#1140)

[vardhans07]

Closes #1140

Context & Fix

Some LLMs serialize omitted or null arguments as literal strings like "null", "none", or "undefined" instead of actual null/None. In list_requests and list_sitemap, we were checking truthiness directly (if scope_id:), which allowed these strings to pass through to the Caido GraphQL query, throwing an Invalid ID format, should be an i32 error.

To fix this, I added a _normalize_optional_id helper in caido_api.py that:

• Treats common string null sentinels and empty whitespace as None.
• Ensures any provided ID parses cleanly into a signed 32-bit integer.
• Normalizes scope_id and parent_id before building the requests.

Also added test cases in tests/test_caido_api.py to cover these sentinel strings, valid numeric IDs, invalid non-integers, and integer overflow.

[greptile-apps]

Greptile Summary

This PR normalizes optional Caido scope and sitemap-parent identifiers, treating common string null sentinels as absent and validating supplied IDs as signed 32-bit integers.

• Applies normalization before request scope filtering and sitemap query selection.
• Adds direct unit coverage for sentinels, numeric normalization, invalid values, and overflow.

Confidence Score: 4/5

The PR appears safe to merge, with a non-blocking test-coverage gap around the production query-building paths.

The normalization and call-site wiring are internally consistent, but the tests exercise only the helper and therefore do not protect the agent-facing filter and sitemap-query behavior.

Files Needing Attention: tests/test_caido_api.py

Important Files Changed

Filename	Overview
strix/tools/proxy/caido_api.py	Adds consistent optional-ID normalization to request filtering and sitemap traversal; no concrete runtime defect was established.
tests/test_caido_api.py	Thoroughly covers the helper itself but omits production call-site behavior such as scope-filter omission and sitemap branch selection.

Prompt To Fix All With AI

### Issue 1
tests/test_caido_api.py:8-9
**Call-site normalization remains untested**

These tests invoke only `_normalize_optional_id`, so they do not verify that sentinel scope IDs omit the request filter or that sentinel parent IDs select the sitemap roots query. Add call-site tests to protect the agent-facing behavior this change is intended to fix.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

_{Reviews (1): Last reviewed commit: "fix(caido): normalize string null and se..." | Re-trigger Greptile}

[vardhans07]

Hi , Just wanted to check in on this. I added those call-site tests a couple of days ago for list_requests_with_client and list_sitemap_with_client to address the bot feedback.

Everything is passing cleanly, so please let me know if this looks good or if you need any other changes. Thank you.