[Klaud Cold] Update qwen3.5-fp8-b200-sglang-agentic-mtp SGLang image to nightly-dev-cu13-20260907-30705c00

[functionstackx]

Summary

Update SGLang image for the Qwen3.5-397B-A17B FP8 B200 AgentX MTP recipe from lmsysorg/sglang:v0.5.16-cu130 (v0.5.16 release) to lmsysorg/sglang:nightly-dev-cu13-20260907-30705c00 (2026-09-07 cu13 dev nightly).

• Tag verified on Docker Hub: last pushed 2026-09-07T01:43:42Z, digest sha256:19b8fa1223cc339c1eae7a5b703f1a8c2543b5b119155bf3d7efaef18f77f007
• Tag commit: sgl-project/sglang@30705c00 ("[Deepseek V4] Keep fp32 routing weights in the mxfp4 trtllm MoE")
• Follows the repo's nightly-dev-cu13-<date>-<hash> convention used by the other B200/B300 SGLang recipes, matching the current cu130 pin.
• Recipe script, NEXTN MTP settings, golden AL 3.39, and the TP4/TP8 concurrency and HiCache grids are unchanged; image-only bump.

Recipes touched: qwen3.5-fp8-b200-sglang-agentic-mtp

Test plan

• full-sweep-enabled sweep passes on cluster:b200-nscale.

🤖 Generated with Claude Code

---

Note

Low Risk
Benchmark recipe pin and changelog only; no runtime code, auth, or serving logic changes.

Overview
Image-only update for the qwen3.5-fp8-b200-sglang-agentic-mtp recipe on cluster:b200-nscale: the SGLang container moves from lmsysorg/sglang:v0.5.16-cu130 to lmsysorg/sglang:nightly-dev-cu13-20260907-30705c00, matching the repo’s nightly-dev-cu13-<date>-<hash> pattern used by other B200 SGLang configs.

perf-changelog.yaml adds an agentic-coding entry for this config key (digest, tag commit, and note that engine flags, NEXTN MTP, golden acceptance length, and TP/HiCache search grids are unchanged).

^{Reviewed by Cursor Bugbot for commit cf0125e. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

Thanks for the contribution! Please reach out to respective companies' CODEOWNER to fill in the latest PR_REVIEW_CHECKLIST.md before pinging core maintainer on Slack for review. In order for the signoff PR check bot to trigger, you must follow the PR_REVIEW_CHECKLIST.md template correctly, including the phrase As a PR reviewer and CODEOWNER, I have reviewed this and have.

For PR verification, add the full-sweep-fail-fast label (strongly recommended) to this PR — the benchmark sweep only runs on labeled PRs. Use full-sweep-enabled only if you need matrix jobs to keep running past a failure.

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. See GitHub's docs on re-running failed jobs

---

感谢你的贡献！请联系相应公司的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，然后再在 Slack 上联系核心维护者进行审阅。为了触发 signoff PR 检查机器人，你必须正确遵循 PR_REVIEW_CHECKLIST.md 模板，包括保留英文语句 As a PR reviewer and CODEOWNER, I have reviewed this and have。

如需进行 PR 验证，请为此 PR 添加 full-sweep-fail-fast 标签（强烈推荐）— 基准测试 sweep 仅在带有标签的 PR 上运行。仅当需要矩阵任务在失败后继续运行时才使用 full-sweep-enabled。

PR 作者有责任确保合并后所有 GitHub Action 任务完全通过。 很多时候失败只是偶发抖动（flake），重新运行失败的任务即可解决。参见 GitHub 关于重新运行失败任务的文档

[github-actions]

Thanks for the contribution! Please reach out to respective companies' CODEOWNER to fill in the latest PR_REVIEW_CHECKLIST.md before pinging core maintainer on Slack for review. In order for the signoff PR check bot to trigger, you must follow the PR_REVIEW_CHECKLIST.md template correctly, including the phrase As a PR reviewer and CODEOWNER, I have reviewed this and have.

For PR verification, add the full-sweep-fail-fast label (strongly recommended) to this PR — the benchmark sweep only runs on labeled PRs. Use full-sweep-enabled only if you need matrix jobs to keep running past a failure.

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. See GitHub's docs on re-running failed jobs

---

感谢你的贡献！请联系相应公司的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，然后再在 Slack 上联系核心维护者进行审阅。为了触发 signoff PR 检查机器人，你必须正确遵循 PR_REVIEW_CHECKLIST.md 模板，包括保留英文语句 As a PR reviewer and CODEOWNER, I have reviewed this and have。

如需进行 PR 验证，请为此 PR 添加 full-sweep-fail-fast 标签（强烈推荐）— 基准测试 sweep 仅在带有标签的 PR 上运行。仅当需要矩阵任务在失败后继续运行时才使用 full-sweep-enabled。

PR 作者有责任确保合并后所有 GitHub Action 任务完全通过。 很多时候失败只是偶发抖动（flake），重新运行失败的任务即可解决。参见 GitHub 关于重新运行失败任务的文档

[github-actions]

Thanks for the contribution! Please reach out to respective companies' CODEOWNER to fill in the latest PR_REVIEW_CHECKLIST.md before pinging core maintainer on Slack for review. In order for the signoff PR check bot to trigger, you must follow the PR_REVIEW_CHECKLIST.md template correctly, including the phrase As a PR reviewer and CODEOWNER, I have reviewed this and have.

For PR verification, add the full-sweep-fail-fast label (strongly recommended) to this PR — the benchmark sweep only runs on labeled PRs. Use full-sweep-enabled only if you need matrix jobs to keep running past a failure.

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. See GitHub's docs on re-running failed jobs

---

感谢你的贡献！请联系相应公司的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，然后再在 Slack 上联系核心维护者进行审阅。为了触发 signoff PR 检查机器人，你必须正确遵循 PR_REVIEW_CHECKLIST.md 模板，包括保留英文语句 As a PR reviewer and CODEOWNER, I have reviewed this and have。

如需进行 PR 验证，请为此 PR 添加 full-sweep-fail-fast 标签（强烈推荐）— 基准测试 sweep 仅在带有标签的 PR 上运行。仅当需要矩阵任务在失败后继续运行时才使用 full-sweep-enabled。

PR 作者有责任确保合并后所有 GitHub Action 任务完全通过。 很多时候失败只是偶发抖动（flake），重新运行失败的任务即可解决。参见 GitHub 关于重新运行失败任务的文档

[claude]

🔴 The appended entry's pr-link: PRLINK_PLACEHOLDER isn't a recognized placeholder, so the automated merge tooling that canonicalizes it into the real PR URL will fail, unlike base where every entry's link is either canonical or the accepted XXX/.../pull/XXX sentinel. Fix: use pr-link: XXX (the only placeholder PR_LINK_PLACEHOLDERS in utils/validate_perf_changelog.py accepts) so prepare_perf_changelog_merge.py canonicalize can replace it automatically.

Extended reasoning...

utils/validate_perf_changelog.py defines PR_LINK_PLACEHOLDERS = {"XXX", ".../pull/XXX"}; validate_added_pr_link() rejects any other value once pr_number is known (line ~156). merge_with_reuse.sh (the flow for full-sweep-enabled/reuse-sweep PRs, matching this PR's label) runs prepare_perf_changelog_merge.py canonicalize at merge time, which calls compare_entries -> validate_added_pr_link on the appended entry's pr-link. Since this entry uses the literal string "PRLINK_PLACEHOLDER" (not "XXX", and not matching docs/configuration-procedures.md's documented "TBD" convention either), canonicalize_appended_links raises ChangelogValidationError("appended entry N has unexpected pr-link 'PRLINK_PLACEHOLDER'"), aborting the scripted merge and forcing a manual fix before this PR can land via that automation.

Verification: normal. The appended entry at perf-changelog.yaml:6931 uses pr-link: PRLINK_PLACEHOLDER, which is not an accepted placeholder. utils/validate_perf_changelog.py:24-27 defines PR_LINK_PLACEHOLDERS = {"XXX", "https://github.com/SemiAnalysisAI/InferenceX/pull/XXX"}. At merge time the full-sweep-enabled flow (utils/merge_with_reuse.sh, set -euo pipefail at line 18) invokes… | normal. The…

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=34168609924
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=34168609924