Feat: Add initial support for v1/messages

[RobGeada]

Description

Using the vLLM conversion between /v1/messages and /v1/chat/completions as a template, this PR adds a similar feature to the NeMo Guardrails server. This allows for the use of the NeMo Guardrails server as the backend endpoint for Claude Code, e.g.:

❯ I've set up some PII filters to auto-redact email addresses, can you try to generate some more?

⏺ FYI, I've run a guardrail.

  Absolutely! Generating realistic test data for PII filters is a smart way to stress-test your regex or NLP-based redaction pipelines. Here
   are a few more sample email addresses, deliberately varied to cover common structural quirks that often challenge naive parsers:

  - <EMAIL_ADDRESS> (utilizes the standard plus-addressing syntax for filtering/tracking)
  - <EMAIL_ADDRESS> (combines an underscore with a multi-part local name)
  - <EMAIL_ADDRESS> (includes a numeric identifier and a nested subdomain structure)
  - <EMAIL_ADDRESS> (uses hyphens in the domain paired with a role-based prefix)
  - <EMAIL_ADDRESS> (mixes a role with a number and a modern application TLD)
  - <EMAIL_ADDRESS> (multiple periods, initials, and a hyphenated domain)
  - <EMAIL_ADDRESS> (longer, department-specific format with a year suffix)

  These are all safely fictional and routed through reserved or common test domains, so they won't attempt real delivery, but they should
  give your PII redaction layer a solid workout across different separators, lengths, casing, and structural edge cases.

  If you're building validation or parsing logic, I'd also recommend throwing in a few intentionally malformed or non-standard variations
  (like user@ or .hidden@domain..com) to ensure your filter gracefully rejects or flags them rather than crashing. Let me know if you need
  them batch-exported in CSV/JSON format, or if you want to test against case-insensitivity, UTF-8/Unicode email quirks, or
  internationalized domain names (IDNs)!

Related Issue(s)

Checklist

• I've read the CONTRIBUTING guidelines.
• I've updated the documentation if applicable.
• I've added tests if applicable.
• @mentions of the person or team responsible for reviewing proposed changes.

[github-advanced-security]

Bandit found more than 20 potential problems in the proposed changes. Check the Files changed tab for more details.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 76.56250% with 75 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (develop@6a3cfd1). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
nemoguardrails/server/anthropic_serving.py	69.58%	66 Missing ⚠️
nemoguardrails/server/schemas/anthropic.py	91.08%	9 Missing ⚠️
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             develop      #42   +/-   ##
==========================================
  Coverage           ?   76.89%           
==========================================
  Files              ?      202           
  Lines              ?    20732           
  Branches           ?        0           
==========================================
  Hits               ?    15941           
  Misses             ?     4791           
  Partials           ?        0           

Flag	Coverage Δ
python	76.89% <76.56%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
nemoguardrails/server/api.py	75.50% <100.00%> (ø)
nemoguardrails/server/schemas/anthropic.py	91.08% <91.08%> (ø)
nemoguardrails/server/anthropic_serving.py	69.58% <69.58%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.