test(agent-observability): pin OTel semconv schema URL to v1.41.0 raw GitHub path

[liustve]

Summary

• Conftest pins raw GitHub path so schema-validation tests are stable when the docs site drifts.

[github-actions]

Claude finished @liustve's task in 1m 39s —— View job

---

Review

Reviewed the diff for PR #774.

• Read prior reviews (none — first pass)
• Read PR diff (conftest.py, +4 / −1)
• Verify usage of _OTEL_SCHEMA_BASE
• Post review

Findings: none — no new substantive issues.

Notes (non-blocking, FYI only):

• The change is well-scoped: pins _OTEL_SCHEMA_BASE to a tagged version (v1.41.0) of the upstream semantic-conventions repo via raw.githubusercontent.com, and the version is extracted to a named constant (_OTEL_SEMCONV_VERSION) so future bumps are a one-line edit. Good.
• validate_otel_genai_schema is consumed by langchain, llama_index, and crewai instrumentor tests for the three schemas (gen-ai-input-messages, gen-ai-output-messages, gen-ai-system-instructions). All three exist under docs/gen-ai/ in the upstream repo at v1.41.0, so the URL substitution should resolve cleanly. The existing _SCHEMA_CACHE keeps repeated calls within a test run cheap.
• The previous URL (https://opentelemetry.io/docs/specs/semconv) is the rendered docs site, not a raw schema host — the pinned raw GitHub URL is the right base for urllib.request.urlopen + json.loads. This likely also fixes silent test failures if the docs site path was returning HTML.
• Tests still require network access on first call (unchanged behavior). Out of scope for this PR, but if flakiness from rate-limited or unreachable raw.githubusercontent becomes an issue, vendoring the three small schema JSONs into the repo is a natural follow-up.

LGTM.
• Branch: di/conftest-pin-semconv-schema