fix: tool schema team_id not effective(#1395)

[caochenlei]

🐛 Bug-fix PR

Before opening this PR please:

1. make lint - passes ruff, mypy, pylint
2. make test - all unit + integration tests green
3. make coverage - ≥ 90 %
4. make docker docker-run-ssl or make podman podman-run-ssl
5. Update relevant documentation.
6. Tested with sqlite and postgres + redis.
7. Manual regression no longer fails. Ensure the UI and /version work correctly.

---

📌 Summary

What problem does this PR fix and why?

fix: tool schema team_id not effective(#1395)

🔁 Reproduction Steps

Link the issue and minimal steps to reproduce the bug.

https://github.com/IBM/mcp-context-forge/issues/1395

🐞 Root Cause

What was wrong and where?

The tool schema team_id was not read when the tool was saved, but instead used parameters from the parameters. It should be read from the schema first, and then from the parameters like other resources

💡 Fix Description

How did you solve it? Key design points.

Prioritize reading team_id and other parameters from the tool schema

🧪 Verification

Check	Command	Status
Lint suite	make lint
Unit tests	make test
Coverage ≥ 90 %	make coverage
Manual regression no longer fails	steps / screenshots

📐 MCP Compliance (if relevant)

• Matches current MCP spec
• No breaking change to MCP clients

✅ Checklist

• Code formatted (make black isort pre-commit)
• No secrets/credentials committed

[crivetimihai]

Closes #1395

[caochenlei]

Closed

[crivetimihai]

Need to better understand the security impact. Under review.