Upgrade llama-stack from v0.6.0 to v0.7.1#2104
Conversation
|
Important Review skippedDraft detected. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
WalkthroughRun configuration YAML files across examples, tests, and root are updated to replace the ChangesRun configuration migration
Estimated code review effort: 2 (Simple) | ~10 minutes Suggested reviewers: 🚥 Pre-merge checks | ✅ 5 | ❌ 1❌ Failed checks (1 inconclusive)
✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
✨ Simplify code
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. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@providers`:
- Line 1: The submodule gitlink for providers points to a commit that is not
available in the providers remote, so CI and fresh clones fail during submodule
fetch. Update the providers pointer to a commit that has already been pushed to
the remote, or push the referenced providers commit first, then refresh the
submodule reference in the parent repo so it resolves cleanly.
In `@run.yaml`:
- Around line 141-145: Update the stale REQUIRED comment near the file_search
tool registration so it no longer mentions llama-stack’s rag-runtime; the
provider is now file-search. Keep the note focused on why this section is needed
for file_search tool calls and ensure the wording matches the current provider
name in the surrounding toolgroup_id/provider_id block.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: 4d81259e-333a-44eb-9e7c-8373f08a3cf3
⛔ Files ignored due to path filters (1)
uv.lockis excluded by!**/*.lock
📒 Files selected for processing (18)
.konflux/requirements.hashes.source.txtexamples/azure-run.yamlexamples/bedrock-run.yamlexamples/run.yamlexamples/vertexai-run.yamlexamples/watsonx-run.yamlproviderspyproject.tomlrun.yamltests/configuration/run.yamltests/e2e-prow/rhoai/configs/run.yamltests/e2e/configs/run-azure.yamltests/e2e/configs/run-bedrock.yamltests/e2e/configs/run-ci.yamltests/e2e/configs/run-rhaiis.yamltests/e2e/configs/run-rhelai.yamltests/e2e/configs/run-vertexai.yamltests/e2e/configs/run-watsonx.yaml
📜 Review details
⏰ Context from checks skipped due to timeout. (8)
- GitHub Check: integration_tests (3.13)
- GitHub Check: integration_tests (3.12)
- GitHub Check: E2E: library mode / ci / group 1
- GitHub Check: E2E: library mode / ci / group 2
- GitHub Check: E2E: library mode / ci / group 3
- GitHub Check: E2E: server mode / ci / group 3
- GitHub Check: E2E: server mode / ci / group 2
- GitHub Check: E2E: server mode / ci / group 1
⚠️ CI failures not shown inline (7)
GitHub Actions: Pyright / 0_Pyright.txt: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Run uv run pyright src
�[36;1muv run pyright src�[0m
shell: /usr/bin/bash -e {0}
env:
UV_PYTHON: 3.12
VIRTUAL_ENV: /home/runner/work/lightspeed-stack/lightspeed-stack/.venv
UV_CACHE_DIR: /home/runner/work/_temp/setup-uv-cache
##[endgroup]
/home/runner/work/lightspeed-stack/lightspeed-stack/src/app/endpoints/responses.py
/home/runner/work/lightspeed-stack/lightspeed-stack/src/app/endpoints/responses.py:968:21 - error: Argument of type "Sequence[OpenAIResponseOutput]" cannot be assigned to parameter "response_items" of type "Sequence[ResponseItem] | None" in function "extract_text_from_response_items"
Type "Sequence[OpenAIResponseOutput]" is not assignable to type "Sequence[ResponseItem] | None"
"Sequence[OpenAIResponseOutput]" is not assignable to "Sequence[ResponseItem]"
Type parameter "_T_co@Sequence" is covariant, but "OpenAIResponseOutput" is not a subtype of "ResponseItem"
Type "OpenAIResponseOutput" is not assignable to type "ResponseItem"
Type "OpenAIResponseOutputMessageReasoningItem" is not assignable to type "ResponseItem"
"Sequence[OpenAIResponseOutput]" is not assignable to "None" (reportArgumentType)
/home/runner/work/lightspeed-stack/lightspeed-stack/src/app/endpoints/responses.py:1104:60 - error: Argument of type "Sequence[OpenAIResponseOutput]" cannot be assigned to parameter "response_items" of type "Sequence[ResponseItem] | None" in function "extract_text_from_response_items"
Type "Sequence[OpenAIResponseOutput]" is not assignable to type "Sequence[ResponseItem] | None"
"Sequence[OpenAIResponseOutput]" is not assignable to "Sequence[ResponseItem]"
Type parameter "_T_co@Sequence" is covariant, but "OpenAIResponseOutput" is not a subtype of "ResponseItem"
Type "OpenAIResponseOutput" is not assignable to type "ResponseItem"
Type "OpenAIResponseOutputMessageReasoningItem" is not assignable to type "ResponseItem"
"Sequen...
GitHub Actions: Pyright / Pyright: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Run uv run pyright src
�[36;1muv run pyright src�[0m
shell: /usr/bin/bash -e {0}
env:
UV_PYTHON: 3.12
VIRTUAL_ENV: /home/runner/work/lightspeed-stack/lightspeed-stack/.venv
UV_CACHE_DIR: /home/runner/work/_temp/setup-uv-cache
##[endgroup]
/home/runner/work/lightspeed-stack/lightspeed-stack/src/app/endpoints/responses.py
/home/runner/work/lightspeed-stack/lightspeed-stack/src/app/endpoints/responses.py:968:21 - error: Argument of type "Sequence[OpenAIResponseOutput]" cannot be assigned to parameter "response_items" of type "Sequence[ResponseItem] | None" in function "extract_text_from_response_items"
Type "Sequence[OpenAIResponseOutput]" is not assignable to type "Sequence[ResponseItem] | None"
"Sequence[OpenAIResponseOutput]" is not assignable to "Sequence[ResponseItem]"
Type parameter "_T_co@Sequence" is covariant, but "OpenAIResponseOutput" is not a subtype of "ResponseItem"
Type "OpenAIResponseOutput" is not assignable to type "ResponseItem"
Type "OpenAIResponseOutputMessageReasoningItem" is not assignable to type "ResponseItem"
"Sequence[OpenAIResponseOutput]" is not assignable to "None" (reportArgumentType)
/home/runner/work/lightspeed-stack/lightspeed-stack/src/app/endpoints/responses.py:1104:60 - error: Argument of type "Sequence[OpenAIResponseOutput]" cannot be assigned to parameter "response_items" of type "Sequence[ResponseItem] | None" in function "extract_text_from_response_items"
Type "Sequence[OpenAIResponseOutput]" is not assignable to type "Sequence[ResponseItem] | None"
"Sequence[OpenAIResponseOutput]" is not assignable to "Sequence[ResponseItem]"
Type parameter "_T_co@Sequence" is covariant, but "OpenAIResponseOutput" is not a subtype of "ResponseItem"
Type "OpenAIResponseOutput" is not assignable to type "ResponseItem"
Type "OpenAIResponseOutputMessageReasoningItem" is not assignable to type "ResponseItem"
"Sequen...
GitHub Actions: Check image building / 0_build-pr.txt: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Fetching submodules
[command]/usr/bin/git submodule sync --recursive
[command]/usr/bin/git -c protocol.version=2 submodule update --init --force --depth=1 --recursive
Submodule 'providers' (https://github.com/lightspeed-core/lightspeed-providers.git) registered for path 'providers'
Cloning into '/home/runner/work/lightspeed-stack/lightspeed-stack/providers'...
##[error]fatal: remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803
GitHub Actions: PR Title Checker / 0_check.txt: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Run thehanimo/pr-title-checker@v1.4.3
with:
GITHUB_***REDACTED***
pass_on_octokit_error: false
configuration_path: .github/pr-title-checker-config.json
##[endgroup]
(node:2171) [DEP0040] DeprecationWarning: The `punycode` module is deprecated. Please use a userland alternative instead.
Using config file .github/pr-title-checker-config.json from repo lightspeed-core/lightspeed-stack [ref: 93e6dc2c5623a0a2c31bf57ac36f91cb0579e0e4]
(Use `node --trace-deprecation ...` to show where the warning was created)
(node:2171) [DEP0169] DeprecationWarning: `url.parse()` behavior is not standardized and prone to errors that have security implications. Use the WHATWG URL API instead. CVEs are not issued for `url.parse()` vulnerabilities.
Creating label (title needs formatting)...
Label (title needs formatting) already created.
Adding label (title needs formatting) to PR...
HttpError: Resource not accessible by integration
##[error]Failed to add label (title needs formatting) to PR
GitHub Actions: PR Title Checker / check: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Run thehanimo/pr-title-checker@v1.4.3
with:
GITHUB_***REDACTED***
pass_on_octokit_error: false
configuration_path: .github/pr-title-checker-config.json
##[endgroup]
(node:2171) [DEP0040] DeprecationWarning: The `punycode` module is deprecated. Please use a userland alternative instead.
Using config file .github/pr-title-checker-config.json from repo lightspeed-core/lightspeed-stack [ref: 93e6dc2c5623a0a2c31bf57ac36f91cb0579e0e4]
(Use `node --trace-deprecation ...` to show where the warning was created)
(node:2171) [DEP0169] DeprecationWarning: `url.parse()` behavior is not standardized and prone to errors that have security implications. Use the WHATWG URL API instead. CVEs are not issued for `url.parse()` vulnerabilities.
Creating label (title needs formatting)...
Label (title needs formatting) already created.
Adding label (title needs formatting) to PR...
HttpError: Resource not accessible by integration
##[error]Failed to add label (title needs formatting) to PR
GitHub Actions: OpenAPI (Spectral) / 0_spectral.txt: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Run set -euo pipefail
�[36;1mset -euo pipefail�[0m
�[36;1muv run python scripts/generate_openapi_schema.py /tmp/openapi-generated.json�[0m
�[36;1mif ! diff -u docs/openapi.json /tmp/openapi-generated.json; then�[0m
�[36;1m echo "::error::docs/openapi.json is out of date. Regenerate with: uv run scripts/generate_openapi_schema.py docs/openapi.json"�[0m
GitHub Actions: OpenAPI (Spectral) / spectral: Upgrade llama-stack from v0.6.0 to v0.7.1
Conclusion: failure
##[group]Run set -euo pipefail
�[36;1mset -euo pipefail�[0m
�[36;1muv run python scripts/generate_openapi_schema.py /tmp/openapi-generated.json�[0m
�[36;1mif ! diff -u docs/openapi.json /tmp/openapi-generated.json; then�[0m
�[36;1m echo "::error::docs/openapi.json is out of date. Regenerate with: uv run scripts/generate_openapi_schema.py docs/openapi.json"�[0m
🧰 Additional context used
🧠 Learnings (2)
📚 Learning: 2026-05-20T08:09:30.641Z
Learnt from: max-svistunov
Repo: lightspeed-core/lightspeed-stack PR: 1580
File: docs/design/llama-stack-config-merge/poc-results/library-mode/synthesized-run.yaml:107-110
Timestamp: 2026-05-20T08:09:30.641Z
Learning: In Llama-stack config YAMLs, when defining a Llama Guard safety shield entry, set `provider_shield_id` to the *guard model identifier* (e.g., `meta-llama/Llama-Guard-3-8B`). Do not use a chat/generative model id (e.g., `openai/gpt-4o-mini`): a chat-model id (or `native_override`) indicates only an override landed and does **not** mean the safety shield is actually gating queries. Ensure any E2E coverage for the related implementation (JIRA/E2E tests) exercises a real Llama Guard model to verify that the shield is effective.
Applied to files:
tests/configuration/run.yamltests/e2e/configs/run-rhelai.yamlexamples/bedrock-run.yamltests/e2e-prow/rhoai/configs/run.yamlexamples/watsonx-run.yamltests/e2e/configs/run-rhaiis.yamltests/e2e/configs/run-azure.yamltests/e2e/configs/run-ci.yamlexamples/vertexai-run.yamltests/e2e/configs/run-watsonx.yamltests/e2e/configs/run-bedrock.yamltests/e2e/configs/run-vertexai.yamlexamples/azure-run.yamlrun.yamlexamples/run.yaml
📚 Learning: 2025-12-18T10:21:03.056Z
Learnt from: are-ces
Repo: lightspeed-core/lightspeed-stack PR: 935
File: run.yaml:114-115
Timestamp: 2025-12-18T10:21:03.056Z
Learning: In run.yaml for Llama Stack 0.3.x, do not add a telemetry provider block under providers. To enable telemetry, set telemetry.enabled: true directly (no provider block required). This pattern applies specifically to run.yaml configuration in this repository.
Applied to files:
run.yaml
🪛 GitHub Actions: Check image building / build-pr
providers
[error] 1-1: Git submodule update failed for submodule 'providers' at 'providers'. fatal: remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803
🪛 GitHub Actions: E2E Tests for Lightspeed Evaluation / 0_E2E Tests for Lightspeed Evaluation job.txt
providers
[error] 1-1: Submodule update failed during 'git submodule update --init --force --depth=1 --recursive' with fatal error: 'remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803'.
🪛 GitHub Actions: E2E Tests for Lightspeed Evaluation / E2E Tests for Lightspeed Evaluation job
providers
[error] 1-1: Submodule update failed during cloning for 'providers'. git submodule update --init --force --depth=1 --recursive returned: "fatal: remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803"
🔇 Additional comments (20)
tests/configuration/run.yaml (2)
20-22: LGTM!Also applies to: 45-46, 88-92, 142-143
1-18: 🎯 Functional CorrectnessCheck the
apislist intests/configuration/run.yaml
Make sure it usesresponses,tool_groups, andconversationsinstead of the oldagentsentry.tests/e2e-prow/rhoai/configs/run.yaml (2)
5-5: LGTM!Also applies to: 14-15, 59-60, 71-71, 80-81, 107-108, 163-164
14-15: 🩺 Stability & AvailabilityEnsure
conversationshas matching storage/provider config.
TheconversationsAPI is enabled here, but this config should also include the correspondingconversationsstorage store and provider entry if the feature is meant to be usable in this environment.tests/e2e/configs/run-azure.yaml (2)
5-5: LGTM!Also applies to: 14-15, 64-71, 80-81, 107-108, 156-157
64-71: 🩺 Stability & AvailabilityConfirm whether
inline::file-searchneedsvector_io
tests/e2e/configs/run-azure.yamlregistersfile-searchbut leavesvector_io: []. Ifinline::file-searchdepends on a vector I/O backend, these E2E configs won’t actually cover file-search behavior.tests/e2e/configs/run-bedrock.yaml (1)
5-5: LGTM!Also applies to: 14-15, 63-69, 78-79, 105-106, 154-155
tests/e2e/configs/run-ci.yaml (1)
5-5: LGTM!Also applies to: 14-15, 58-64, 73-74, 100-101, 145-146
tests/e2e/configs/run-rhaiis.yaml (1)
5-5: LGTM!Also applies to: 14-15, 65-71, 80-81, 107-108, 156-157
tests/e2e/configs/run-rhelai.yaml (1)
5-5: LGTM!Also applies to: 14-15, 65-71, 80-81, 107-108, 156-157
tests/e2e/configs/run-vertexai.yaml (1)
5-5: LGTM!Also applies to: 14-15, 64-70, 79-80, 106-107, 151-152
tests/e2e/configs/run-watsonx.yaml (1)
5-5: LGTM!Also applies to: 14-15, 65-71, 80-81, 107-108, 152-153
pyproject.toml (1)
31-33: LGTM!run.yaml (2)
4-4: LGTM!Also applies to: 13-14, 57-66, 75-76, 102-103
13-14: Checkconversationsstorage beside everyconversationsAPI entry. If any otherrun.yamlin this PR addsconversationsunderapis, it also needs a matchingstorage.stores.conversationsentry to avoid persistence failures.examples/run.yaml (1)
10-10: LGTM!Also applies to: 19-20, 64-69, 78-79, 105-106, 147-148
examples/azure-run.yaml (1)
5-5: LGTM!Also applies to: 14-15, 56-57, 68-68, 77-78, 104-105, 160-161
examples/bedrock-run.yaml (1)
4-4: LGTM!Also applies to: 13-14, 62-63, 74-74, 83-84, 110-111, 159-160
examples/vertexai-run.yaml (1)
4-15: LGTM!Also applies to: 55-57, 68-78, 101-105, 156-157
examples/watsonx-run.yaml (1)
4-4: LGTM!Also applies to: 13-14, 64-65, 76-76, 85-86, 112-113, 168-169
| @@ -1 +1 @@ | |||
| Subproject commit 9ab872996b1b95448a9332bc3b468fe27b4ee2e6 | |||
| Subproject commit 537fab9854b1b6013bab30d0f295685979c2b803 | |||
There was a problem hiding this comment.
🩺 Stability & Availability | 🔴 Critical | 🏗️ Heavy lift
Submodule points to a commit that is not present in the providers remote — CI is broken.
The pointer references 537fab9854b1b6013bab30d0f295685979c2b803, but every submodule fetch fails with upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803. This matches the PR note that the corresponding providers change still has to be made and pulled in. The commit must be pushed to the providers remote (or this pointer must be set to an already-published commit) before merge; otherwise all clones and CI jobs that init submodules will fail.
🧰 Tools
🪛 GitHub Actions: Check image building / build-pr
[error] 1-1: Git submodule update failed for submodule 'providers' at 'providers'. fatal: remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803
🪛 GitHub Actions: E2E Tests for Lightspeed Evaluation / 0_E2E Tests for Lightspeed Evaluation job.txt
[error] 1-1: Submodule update failed during 'git submodule update --init --force --depth=1 --recursive' with fatal error: 'remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803'.
🪛 GitHub Actions: E2E Tests for Lightspeed Evaluation / E2E Tests for Lightspeed Evaluation job
[error] 1-1: Submodule update failed during cloning for 'providers'. git submodule update --init --force --depth=1 --recursive returned: "fatal: remote error: upload-pack: not our ref 537fab9854b1b6013bab30d0f295685979c2b803"
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@providers` at line 1, The submodule gitlink for providers points to a commit
that is not available in the providers remote, so CI and fresh clones fail
during submodule fetch. Update the providers pointer to a commit that has
already been pushed to the remote, or push the referenced providers commit
first, then refresh the submodule reference in the parent repo so it resolves
cleanly.
Source: Pipeline failures
| - toolgroup_id: builtin::file_search # Register the RAG tool | ||
| provider_id: file-search | ||
| # REQUIRED: This section is necessary for file_search tool calls to work. | ||
| # Without it, llama-stack's rag-runtime silently fails all file_search operations | ||
| # with no error logged. |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Update stale comment referencing rag-runtime.
The comment at lines 143-145 still references llama-stack's rag-runtime, but the provider was renamed to file-search in this PR. This could mislead anyone debugging file_search issues into looking for a non-existent rag-runtime provider.
📝 Proposed fix for stale comment
tool_groups:
- toolgroup_id: builtin::file_search # Register the RAG tool
provider_id: file-search
# REQUIRED: This section is necessary for file_search tool calls to work.
-# Without it, llama-stack's rag-runtime silently fails all file_search operations
+# Without it, llama-stack's file-search provider silently fails all file_search operations
# with no error logged.📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| - toolgroup_id: builtin::file_search # Register the RAG tool | |
| provider_id: file-search | |
| # REQUIRED: This section is necessary for file_search tool calls to work. | |
| # Without it, llama-stack's rag-runtime silently fails all file_search operations | |
| # with no error logged. | |
| tool_groups: | |
| - toolgroup_id: builtin::file_search # Register the RAG tool | |
| provider_id: file-search | |
| # REQUIRED: This section is necessary for file_search tool calls to work. | |
| # Without it, llama-stack's file-search provider silently fails all file_search operations | |
| # with no error logged. |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@run.yaml` around lines 141 - 145, Update the stale REQUIRED comment near the
file_search tool registration so it no longer mentions llama-stack’s
rag-runtime; the provider is now file-search. Keep the note focused on why this
section is needed for file_search tool calls and ensure the wording matches the
current provider name in the surrounding toolgroup_id/provider_id block.
Breaking changes addressed (from llama-stack v0.7.0 release): 1. APIs section updates: - agents → responses - Added tool_groups and conversations APIs 2. Provider section renames: - agents: → responses: - Provider IDs: meta-reference → builtin - Provider types: inline::meta-reference → inline::builtin 3. RAG refactor: - Provider: inline::rag-runtime → inline::file-search - Provider IDs: rag-runtime → file-search - Toolgroup IDs: builtin::rag → builtin::file_search 4. Dependencies: - llama-stack: 0.6.0 → 0.7.1 - llama-stack-client: 0.6.0 → 0.7.1 - llama-stack-api: 0.6.0 → 0.7.1 5. Version compatibility: - Updated MAXIMAL_SUPPORTED_LLAMA_STACK_VERSION to 0.7.1 6. Regenerated dependency hashes Files updated: - pyproject.toml, uv.lock - run.yaml (root) - examples/*.yaml (5 files) - tests/configuration/run.yaml - tests/e2e/configs/run-*.yaml (7 files) - tests/e2e-prow/rhoai/configs/run.yaml - .konflux/requirements.hashes.source.txt - src/constants.py Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Updates providers submodule reference to commit 537fab9 which includes run.yaml configuration updates for llama-stack v0.7.1 compatibility. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
anik120
force-pushed
the
llama-stack-0.7.1
branch
from
5d5cb96 to
d8e00f7
Compare
llama-stack v0.7.1 removed the client.toolgroups API. Toolgroups are now auto-registered from run.yaml provider specs. Code changes: - Remove register_mcp_servers_async() and _register_mcp_toolgroups_async() - Remove call to register_mcp_servers_async() from startup - Refactor tools endpoint to use HTTP GET /v1/tools instead of client.toolgroups - Add _normalize_and_enrich_tool_dict() helper - Add ReasoningItem to ResponseItem union - Handle list-type function call outputs in conversations Test changes: - Remove test_mcp_servers.py (dynamic registration no longer supported) - Remove test_common.py (register_mcp_servers_async removed) - Rewrite test_tools.py to mock HTTP API instead of client.toolgroups - Update e2e llama_stack_utils.py with no-op stub for backward compat Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
llama-stack v0.7.1 doesn't support reasoning.encrypted_content in the include parameter. pydantic-ai automatically adds it based on model profile flags, causing 400 Bad Request errors. Additionally, the Responses API uses max_output_tokens instead of max_tokens (different from the Chat Completions API). Changes: - Remove reasoning.encrypted_content from IncludeParameter type - Override _responses_create in LlamaStackResponsesModel to: - Skip adding reasoning.encrypted_content to include list - Map max_tokens -> max_output_tokens for Responses API This fixes runtime query failures after the v0.7.1 upgrade. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
anik120
force-pushed
the
llama-stack-0.7.1
branch
from
d8e00f7 to
7046dce
Compare
|
Feel like this should rebase to main when #1931 merged. A lot of the dependencies it needs for llama-stack-0.7.1 can be provided by RHOAI 3.5 index. Note that |
2 participants
Description
Breaking changes addressed (from llama-stack v0.7.0 release):
APIs section updates:
Provider section renames:
RAG refactor:
Dependencies:
Regenerated dependency hashes
Files updated:
Co-Authored-By: Claude Sonnet 4.5 noreply@anthropic.com
Type of change
Tools used to create PR
Identify any AI code assistants used in this PR (for transparency and review context)
Related Tickets & Documents
Checklist before requesting a review
Testing
Summary by CodeRabbit
New Features
Chores