LCORE-1285: Update Llama Stack to 0.5.2

[jrobertboos]

Description

Updated Llama Stack to 0.5.X in order to enable the network configuration on providers so that TLS and Proxy support can be added.

Type of change

• Refactor
• New feature
• Bug fix
• CVE fix
• Optimization
• Documentation Update
• Configuration Update
• Bump-up service version
• Bump-up dependent library
• Bump-up library or tool used for development (does not change the final image)
• CI configuration change
• Konflux configuration change
• Unit tests improvement
• Integration tests improvement
• End to end tests improvement
• Benchmarks improvement

Tools used to create PR

Identify any AI code assistants used in this PR (for transparency and review context)

• Assisted-by: N/A
• Generated by: N/A

Related Tickets & Documents

• Related Issue: LCORE-1250, LCORE-1252
• Closes: LCORE-1285

Checklist before requesting a review

• I have performed a self-review of my code.
• PR has passed all pre-merge test jobs.
• If it is a core feature, I have added thorough tests.

Testing

• Please provide detailed steps to perform tests related to this code change.
• How were the fix/results from this change verified? Please provide relevant screenshots or results.

Summary by CodeRabbit

• Chores

  • Updated llama-stack packages to v0.5.2 and bumped maximum supported Llama Stack version.
  • Updated build/requirements pins and hashes, and added/adjusted several build-related dependency entries.

• Bug Fixes

  • Improved robustness of content-response handling to avoid attribute access errors.

• Tests

  • Updated end-to-end test expectations to reflect the new llama-stack version.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 9d0bc77c-f111-45fa-a402-605764bd9752

📥 Commits

Reviewing files that changed from the base of the PR and between ef914b0 and 280741f.

⛔ Files ignored due to path filters (1)

• uv.lock is excluded by !**/*.lock

📒 Files selected for processing (3)

• pyproject.toml
• requirements-build.txt
• requirements.hashes.source.txt

🚧 Files skipped from review as they are similar to previous changes (1)

• pyproject.toml

---

Walkthrough

Updates llama-stack packages and related version constant to 0.5.2, adds/updates multiple build and requirements entries (including hashes), and makes defensive attribute access changes in response parsing; test expectation updated to match the new llama-stack version.

Changes

Cohort / File(s)	Summary
Dependency & Build Pins pyproject.toml, requirements-build.txt, requirements.hashes.source.txt	Bumped llama-stack packages to 0.5.2 and updated many dependency pins and SHA256 hashes; added/adjusted build-related entries (e.g., cython, maturin) and commented optional deps.
Version Constant & Tests src/constants.py, tests/e2e/features/info.feature	Updated MAXIMAL_SUPPORTED_LLAMA_STACK_VERSION constant and test assertion to reflect llama-stack 0.5.2.
Response Parsing Robustness src/utils/responses.py	Replaced direct part.type access with getattr(part, "type", None) in multiple locations to avoid AttributeError for parts missing a type attribute; behavior otherwise unchanged.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and specifically summarizes the main change: updating Llama Stack dependencies to version 0.5.2, which aligns with all file modifications.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

src/utils/responses.py (1)

1142-1151: Log skipped content parts on the fallback path.

This avoids the crash, but it also makes unsupported parts disappear silently. A small debug log here would make empty extracted text much easier to diagnose.

💡 Suggested change

     for part in content:
         part_type = getattr(part, "type", None)
         if part_type == "input_text":
             input_text_part = cast(InputTextPart, part)
             if input_text_part.text:
                 text_fragments.append(input_text_part.text.strip())
         elif part_type == "output_text":
             output_text_part = cast(OutputTextPart, part)
             if output_text_part.text:
                 text_fragments.append(output_text_part.text.strip())
         elif part_type == "refusal":
             refusal_part = cast(ContentPartRefusal, part)
             if refusal_part.refusal:
                 text_fragments.append(refusal_part.refusal.strip())
+        else:
+            logger.debug("Skipping unsupported response content part: %r", part)

As per coding guidelines, "Use logger.debug() for detailed diagnostic information".

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/utils/responses.py` around lines 1142 - 1151, The fallback path that
skips unsupported content parts currently silently drops them; update the logic
around the part_type checks (look for the variables part_type, part, and the
text_fragments list in the function handling response parts) to call
logger.debug() when encountering non-text/unsupported part types (e.g., in the
final elif/else branch where "refusal" or other types are handled) so the debug
log includes the part_type and a small representation of the part (or its id) to
aid diagnosis; ensure you use the module's existing logger (or import one if
missing) and keep the message concise and diagnostic.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@src/utils/responses.py`:
- Around line 1142-1151: The fallback path that skips unsupported content parts
currently silently drops them; update the logic around the part_type checks
(look for the variables part_type, part, and the text_fragments list in the
function handling response parts) to call logger.debug() when encountering
non-text/unsupported part types (e.g., in the final elif/else branch where
"refusal" or other types are handled) so the debug log includes the part_type
and a small representation of the part (or its id) to aid diagnosis; ensure you
use the module's existing logger (or import one if missing) and keep the message
concise and diagnostic.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: d703aad5-1415-4222-9323-04cdd2ee2b0c

📥 Commits

Reviewing files that changed from the base of the PR and between e73778a and ef914b0.

⛔ Files ignored due to path filters (1)

• uv.lock is excluded by !**/*.lock

📒 Files selected for processing (4)

• pyproject.toml
• src/constants.py
• src/utils/responses.py
• tests/e2e/features/info.feature

[asimurka]

LGTM

[tisnik]

Please try to update Konfux references as well. This way we'll see if/how it builds in Konflux

[jrobertboos]

Please try to update Konfux references as well. This way we'll see if/how it builds in Konflux

@tisnik By this do you mean running make konflux-requirements or is there something else that I would need to do?

[tisnik]

LGTM