superseded by #1263: make Azure and cross-provider fallbacks executable by seonghobae · Pull Request #1213 · ContextualWisdomLab/.github

seonghobae · 2026-08-22T00:05:08Z

Superseded by #1263 and closed without merge.

The effective eight-file patch moved unchanged to protected main at exact head f89c0c58ca3f31611af0d2d58ec9c8acbb70b61f; the stable patch-id is identical. This PR's unprotected stack base forced pull_request_target Strix to execute the older trusted gate, so it could only reproduce the openai-direct/ alias failure that the patch fixes. #1263 removes that circular validation dependency without losing product code, weakening checks, rewriting history, or bypassing protection.

coderabbitai · 2026-08-22T00:05:17Z

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 473c553a-ae25-4304-9280-0607d461d10c

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

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

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

seonghobae · 2026-08-22T00:05:36Z

@opencode-agent Review only exact current head 5dbf903. Verify the trusted status-3 boundary, unchanged-finding attribution, provider-exhaustion neutralization, and fail-closed changed/unmapped/configuration paths. Submit a formal Reviews API verdict without changing the branch, self-approving, or bypassing protection.

seonghobae · 2026-08-22T00:21:37Z

@opencode-agent Re-review exact current head 5dbf903 now that all hosted checks are terminal-success and the informational Devin thread is resolved. Submit the formal Reviews API verdict; do not mutate the branch or bypass protection.

seonghobae · 2026-08-22T00:38:39Z

@opencode-agent Please review exact head d065e69fa48fbf96251433f8be77934ccf03b159. The follow-up narrowly recognizes the observed scanner-local RuntimeError: loginAsGuest ... curl exit 7 ... 127.0.0.1 port 48080 Caido bootstrap outage only when no positive vulnerability/severity signal exists; all other runtime failures and any finding remain fail-closed. Local evidence: 1,333 Python tests passed, 1 skipped, 16 subtests passed; bash scripts/ci/test_strix_quick_gate.sh passed; targeted workflow classifier tests (13) and actionlint passed. Please provide an independent exact-head review.

seonghobae

Revalidated exact head 6b03822 after merging protected main 634a41a: 1333 passed, 1 skipped, 16 subtests; 100% statement and branch coverage; 100% production docstrings; focused Strix shell, Bash syntax, and workflow validation passed. Existing review threads remain resolved.

…ovider-exhaustion

seonghobae · 2026-08-22T08:09:25Z

Exact-head verification refresh

Exact head: d1a5a1e
Exact base: 5a8b837
Full Python suite: 1,352 passed, 1 skipped, 16 subtests passed
Coverage: 100% statements and branches, 8,318 statements and 3,208 branches
Production docstrings: 740/740
Full shell regression suite: test_strix_quick_gate PASS
Compilation, bash syntax, and diff checks: PASS
Observed regression: the exact Azure LiteLLM unsupported temperature 0.2 plus missing gpt-5.6-sol fallback-group error crosses the existing distinct-model fallback boundary; split-line marker assembly remains non-retryable and findings remain fail-closed.

coderabbitai · 2026-08-22T09:23:26Z

Caution

Review failed

An error occurred during the review process. Please try again later.

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

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

…ack' into HEAD

…ustion' into HEAD

…edentials run_strix_once() computed normalized_model only for credential/API-base selection (is_vertex_model, is_github_models_model, and the openai_direct/* fallback-key case), but called child_model_for_api_base() with the raw, un-normalized $model. A hyphenated STRIX_FALLBACK_MODELS entry (openai-direct/gpt-5.6-luna) therefore reached child_model_for_api_base unrewritten -- its own openai_direct/* case never matched a hyphen -- and was dispatched to LiteLLM verbatim as an unrecognized provider string, even though normalize_model() (already fixed in this branch) and child_model_for_api_base() both handle the alias correctly on their own. Observed live on this PR's own strix check: the run advanced through two NVIDIA NIM rate limits to the openai-direct/gpt-5.6-luna fallback, the log recorded "model=openai-direct/gpt-5.6-luna" verbatim, and litellm.BadRequestError: LLM Provider NOT provided ended the scan. Move the normalized_model computation before the child_model_for_api_base call and reuse it there, instead of computing it a second time afterward for credentials only. New regression test extracts run_strix_once's own two-statement model/API-base resolution call site verbatim (rather than reimplementing the composition) so a future call site edit that stops normalizing before dispatch fails this test, not just live NVIDIA NIM traffic.

Refresh the gap baseline against the current open queue after #1253-#1256 closed unmerged, record #1257/#1213 head motion, and correct CWL-MASTER-CONTEXT so naruon#975 is Done rather than In Progress. The SHA table remains a snapshot, not merge authorization.

devin-ai-integration

Devin Review found 2 new potential issues.

seonghobae · 2026-08-23T14:26:49Z

Exact-current-head verification for 8d0f7ca88f6e0c9db8e2363721fbeaa59acebd6a:

The causal predecessor 5ab5834 normalizes openai-direct/* before LiteLLM dispatch; the focused NVIDIA/fallback suite passes 16 tests.
Live canary fix(governance): preserve proposal branch create transition #1176 run https://github.com/ContextualWisdomLab/.github/actions/runs/32638743224, job 97192405871, exhausted NVIDIA NIM with HTTP 429 and then passed raw openai-direct/gpt-5.6-luna to LiteLLM, which ended with LLM Provider NOT provided. This is the exact dispatch defect repaired by 5ab5834.
Live canary fix(opencode): split review surfaces, give NIM two hours, and remove GitHub Models #1052 run https://github.com/ContextualWisdomLab/.github/actions/runs/32641474729, job 97199115489, completed an authoritative assessment with Vulnerabilities 0 and a source-aware report, but protected-main classification rejected the exact MODEL QUALITY WARNING. The complete-line classifier and malicious-suffix controls retained on this head repair that false negative.
The current-head report warning repair accepts only the complete timestamped strix.tools.web_search.tool: web_search invoked without PERPLEXITY_API_KEY configured advisory. Its exact positive fixture passes, while the malicious-suffix negative control remains fail-closed.
Full local central Python suite: 1397 passed, 1 skipped, 16 subtests passed.
Full local scripts/ci/test_strix_quick_gate.sh: PASS.
Hosted run https://github.com/ContextualWisdomLab/.github/actions/runs/32644638359, job 97206910172, checked out exact SHA 8d0f7ca88f6e0c9db8e2363721fbeaa59acebd6a and passed its 8 focused toolchain contracts (1 skipped).
Both current review threads were verified as informational against the exact boundary above and resolved; unresolved threads are now 0.

This is source/test evidence only. The PR still targets stacked base fix/organization-loop-oidc-fallback@dfb8e261c81705841111dd4ad1712a9fb6c767d1; no main-target full hosted Strix acceptance or exact-current-head formal Reviews API verdict is claimed. Please issue a substantive formal verdict for this exact head once the stack supplies the required hosted evidence.

Snapshot the current open-PR inventory after #1262/#1213 closed unmerged as superseded. Record #1263 as the G-03 main-line successor. This is not merge authorization.

fix: separate Strix baseline provider exhaustion

5dbf903

seonghobae added this to naruon Platform Roadmap Aug 22, 2026

github-project-automation Bot moved this to Todo in naruon Platform Roadmap Aug 22, 2026

seonghobae moved this from Todo to In Progress in naruon Platform Roadmap Aug 22, 2026

seonghobae enabled auto-merge (squash) August 22, 2026 00:05

This comment was marked as resolved.

Sign in to view

opencode-agent Bot disabled auto-merge August 22, 2026 00:08

fix(strix): neutralize empty Caido bootstrap outages

d065e69

This comment was marked as resolved.

Sign in to view

seonghobae mentioned this pull request Aug 22, 2026

fix(coverage): admit exact base VCS sources #1210

Merged

Merge branch 'main' into fix/strix-baseline-provider-exhaustion

b9dd912

This comment was marked as resolved.

Sign in to view

Merge latest main into Strix provider fallback branch

6b03822

seonghobae commented Aug 22, 2026

View reviewed changes

This comment was marked as resolved.

Sign in to view

fix(strix): preserve primary-only baseline result

7179abb

This comment was marked as resolved.

Sign in to view

refactor(strix): remove obsolete fallback probe

94eafa4

This comment was marked as resolved.

Sign in to view

seonghobae added 3 commits August 22, 2026 16:40

fix(strix): fallback on unsupported sampling defaults

51a4aa7

Merge remote-tracking branch 'origin/main' into fix/strix-baseline-pr…

b1ab66f

…ovider-exhaustion

Merge remote-tracking branch 'origin/main' into fix/strix-baseline-pr…

d1a5a1e

…ovider-exhaustion

This comment was marked as resolved.

Sign in to view

Merge current main into Strix provider exhaustion fix

8646703

seonghobae added 2 commits August 23, 2026 20:20

Merge parent and repair Azure sampling fallback routing

83c2944

Merge remote-tracking branch 'origin/fix/organization-loop-oidc-fallb…

6f6e507

…ack' into HEAD

seonghobae changed the title ~~fix: preserve Strix PR scope on provider exhaustion~~ fix(strix): route unsupported Azure sampling failures to a valid fallback Aug 23, 2026

fix(strix): accept clean fallback model advisory

a675dac

This comment was marked as resolved.

Sign in to view

seonghobae added 3 commits August 23, 2026 21:15

fix(strix): accept clean HF advisory

32543e0

docs(strix): record clean fallback advisories

86408ba

fix(strix): preserve fail-closed warning scan

9dcd1d6

This comment was marked as resolved.

Sign in to view

seonghobae added 3 commits August 23, 2026 21:29

fix(strix): isolate cross-provider OpenAI fallback transport

3868a21

Merge remote-tracking branch 'origin/fix/strix-baseline-provider-exha…

6493023

…ustion' into HEAD

fix(strix): bound clean advisory filtering

5edaf36

seonghobae changed the title ~~fix(strix): route unsupported Azure sampling failures to a valid fallback~~ fix(strix): make Azure and cross-provider fallbacks executable Aug 23, 2026

This was referenced Aug 23, 2026

fix(strix): recognize the hyphenated openai-direct fallback alias #1256

Closed

fix(strix): preserve NIM fallback evidence #1254

Closed

fix(strix): accept complete scans with benign model-quality advisory #1255

Closed

This comment was marked as resolved.

Sign in to view

seonghobae mentioned this pull request Aug 23, 2026

docs: refresh live product and technical gap baseline #1252

Merged

This comment was marked as resolved.

Sign in to view

fix(strix): ignore exact optional web search advisory

8d0f7ca

devin-ai-integration Bot reviewed Aug 23, 2026

View reviewed changes

Comment thread scripts/ci/strix_quick_gate.sh

Comment thread scripts/ci/strix_quick_gate.sh

This was referenced Aug 23, 2026

superseded by #1263: rewrite openai-direct fallback for LiteLLM #1262

Closed

fix(strix): make Azure and cross-provider fallbacks executable #1263

Open

seonghobae changed the title ~~fix(strix): make Azure and cross-provider fallbacks executable~~ superseded by #1263: make Azure and cross-provider fallbacks executable Aug 23, 2026

seonghobae closed this Aug 23, 2026

github-project-automation Bot moved this from In Progress to Done in naruon Platform Roadmap Aug 23, 2026

Uh oh!

Conversation

seonghobae commented Aug 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

coderabbitai Bot commented Aug 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

seonghobae commented Aug 22, 2026

Uh oh!

This comment was marked as resolved.

Uh oh!

seonghobae commented Aug 22, 2026

Uh oh!

seonghobae commented Aug 22, 2026

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

seonghobae left a comment

Choose a reason for hiding this comment

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

seonghobae commented Aug 22, 2026

Exact-head verification refresh

Uh oh!

This comment was marked as resolved.

Uh oh!

coderabbitai Bot commented Aug 22, 2026

Review failed

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

This comment was marked as resolved.

Uh oh!

devin-ai-integration Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

seonghobae commented Aug 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

seonghobae commented Aug 22, 2026 •

edited

Loading

coderabbitai Bot commented Aug 22, 2026 •

edited

Loading

seonghobae commented Aug 23, 2026 •

edited

Loading