Organize benchmarks into weekly and special runs

[aantn]

• other minor changes like changing the title in each benchmark to be the date of its run, so pages are listed more nicely in the docs sidebar.

[coderabbitai]

Walkthrough

Reorganizes benchmark history into weekly and special categories, updates workflow and local script to write timestamped reports under weekly/, adds corresponding docs and navigation files, and revises titles/metadata in specific result pages. Minor edit in running-evals documentation.

Changes

Cohort / File(s)	Summary of Changes
CI workflow output path .github/workflows/eval-benchmarks.yaml	Update history output path to docs/development/evaluations/history/weekly/results_${TIMESTAMP}.md in the Generate benchmark report step.
History docs root nav + index docs/development/evaluations/history/.nav.yml, docs/development/evaluations/history/index.md	Replace wildcard nav with explicit sections: Weekly and Special; index updated to describe weekly runs and introduce special runs.
Weekly docs docs/development/evaluations/history/weekly/.nav.yml, docs/development/evaluations/history/weekly/index.md, docs/development/evaluations/history/weekly/results_20251012_170303.md	Add weekly nav (desc sort, index + all items) and index; adjust a weekly results page title to dated form and add metadata lines (duration, iterations, judge model).
Special docs docs/development/evaluations/history/special/.nav.yml, docs/development/evaluations/history/special/index.md, docs/development/evaluations/history/special/results_20250930_153753.md, docs/development/evaluations/history/special/results_20251008_053744.md	Add special nav (desc sort, index + wildcard) and index; update two result pages’ titles/metadata.
Eval docs wording docs/development/evaluations/running-evals.md	Remove caveat about rare mock cases; simplify sentence about mock data capturing one path.
Local benchmark script run_benchmarks_local.sh	Generate timestamped history under weekly/ (ensures dir, constructs filename, runs report); remove old non-weekly copy step; update finalization to reference new history file and its line count; minor control-flow adjustment.

Sequence Diagram(s)

sequenceDiagram
  autonumber
  participant Dev as Developer
  participant Script as run_benchmarks_local.sh
  participant Gen as report_generator
  participant FS as docs/.../history/weekly/

  Dev->>Script: Run local benchmarks
  Script->>Script: Ensure weekly/ dir exists
  Script->>Gen: Generate latest results
  Script->>Gen: Generate timestamped history report
  Gen-->>FS: Write weekly/results_${TIMESTAMP}.md
  Script-->>Dev: Log history file path and line count

Loading

sequenceDiagram
  autonumber
  participant CI as GitHub Actions
  participant Step as Generate benchmark report
  participant FS as docs/.../history/weekly/

  CI->>Step: Execute workflow job
  Step-->>FS: Write weekly/results_${TIMESTAMP}.md

Loading

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• Add benchmarks for Claude 4.5 #1017: Earlier change adding timestamped history generation and docs/nav adjustments; this PR moves those outputs into weekly/ and expands nav.
• Make benchmarks more prominent in docs #1030: Updates to benchmarks documentation and history pages similar in scope (nav, headers, metadata).

Suggested reviewers

• Sheeproid
• arikalon1

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title Check	✅ Passed	The title concisely captures the main change of reorganizing benchmark reports into separate weekly and special runs and directly reflects the modifications to documentation and workflow logic described in the PR. It focuses on the primary objective without extraneous details and uses clear, specific language. This enables a teammate scanning history to immediately understand the purpose of the changeset. Therefore, it accurately summarizes the core update.
Description Check	✅ Passed	The description highlights the update of benchmark titles to use their run dates, which aligns with the documented changes to the navigation sidebar and metadata in the PR. It is directly related to the changeset and reflects actual modifications made in the documentation. Although brief, it remains on-topic and informs about a visible aspect of the update.
Docstring Coverage	✅ Passed	No functions found in the changes. Docstring coverage check skipped.

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch better-benchmarks-organization

---

📜 Recent review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between b65307d and c2fb63d.

📒 Files selected for processing (12)

• .github/workflows/eval-benchmarks.yaml (1 hunks)
• docs/development/evaluations/history/.nav.yml (1 hunks)
• docs/development/evaluations/history/index.md (1 hunks)
• docs/development/evaluations/history/special/.nav.yml (1 hunks)
• docs/development/evaluations/history/special/index.md (1 hunks)
• docs/development/evaluations/history/special/results_20250930_153753.md (1 hunks)
• docs/development/evaluations/history/special/results_20251008_053744.md (1 hunks)
• docs/development/evaluations/history/weekly/.nav.yml (1 hunks)
• docs/development/evaluations/history/weekly/index.md (1 hunks)
• docs/development/evaluations/history/weekly/results_20251012_170303.md (1 hunks)
• docs/development/evaluations/running-evals.md (1 hunks)
• run_benchmarks_local.sh (2 hunks)

👮 Files not reviewed due to content moderation or server errors (3)

• docs/development/evaluations/history/special/results_20250930_153753.md
• docs/development/evaluations/history/special/results_20251008_053744.md
• docs/development/evaluations/history/weekly/results_20251012_170303.md

🧰 Additional context used

📓 Path-based instructions (1)

docs/**/*.md

📄 CodeRabbit inference engine (CLAUDE.md)

In MkDocs content, always add a blank line between a header or bold text and a following list so lists render correctly

Files:

• docs/development/evaluations/running-evals.md
• docs/development/evaluations/history/special/results_20250930_153753.md
• docs/development/evaluations/history/weekly/results_20251012_170303.md
• docs/development/evaluations/history/special/results_20251008_053744.md
• docs/development/evaluations/history/index.md
• docs/development/evaluations/history/special/index.md
• docs/development/evaluations/history/weekly/index.md

🧠 Learnings (1)

📚 Learning: 2025-08-05T06:14:39.523Z

Learnt from: aantn
PR: robusta-dev/holmesgpt#783
File: tests/llm/fixtures/test_ask_holmes/100_historical_logs/payment-api.yaml:49-70
Timestamp: 2025-08-05T06:14:39.523Z
Learning: For evaluation test fixtures in the holmesgpt project, security contexts and security hardening are not priorities. The focus should be on functionality and test reliability rather than adding security configurations to Kubernetes manifests used in evals.

Applied to files:

• docs/development/evaluations/running-evals.md

🪛 LanguageTool

docs/development/evaluations/history/index.md

[grammar] ~13-~13: There might be a mistake here.
Context: ...such as: - Comparing self-hosted models - Testing new model versions - Performance...

(QB_NEW_EN)

---

[grammar] ~14-~14: There might be a mistake here.
Context: ...sted models - Testing new model versions - Performance analysis for specific scenar...

(QB_NEW_EN)

---

[grammar] ~15-~15: There might be a mistake here.
Context: ...formance analysis for specific scenarios - Custom model comparisons See the **Spec...

(QB_NEW_EN)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (3)

• GitHub Check: llm_evals
• GitHub Check: Pre-commit checks
• GitHub Check: build

---

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.}

[github-actions]

Results of HolmesGPT evals

• ask_holmes: 34/35 test cases were successful, 0 regressions, 1 setup failures

Test suite	Test case	Status
ask	01_how_many_pods	✅
ask	02_what_is_wrong_with_pod	✅
ask	04_related_k8s_events	✅
ask	05_image_version	✅
ask	09_crashpod	✅
ask	10_image_pull_backoff	✅
ask	110_k8s_events_image_pull	✅
ask	11_init_containers	✅
ask	13a_pending_node_selector_basic	✅
ask	14_pending_resources	✅
ask	15_failed_readiness_probe	✅
ask	17_oom_kill	✅
ask	19_detect_missing_app_details	✅
ask	20_long_log_file_search	✅
ask	24_misconfigured_pvc	✅
ask	24a_misconfigured_pvc_basic	✅
ask	28_permissions_error	🚧
ask	39_failed_toolset	✅
ask	41_setup_argo	✅
ask	42_dns_issues_steps_new_tools	✅
ask	43_current_datetime_from_prompt	✅
ask	45_fetch_deployment_logs_simple	✅
ask	51_logs_summarize_errors	✅
ask	53_logs_find_term	✅
ask	54_not_truncated_when_getting_pods	✅
ask	59_label_based_counting	✅
ask	60_count_less_than	✅
ask	61_exact_match_counting	✅
ask	63_fetch_error_logs_no_errors	✅
ask	79_configmap_mount_issue	✅
ask	83_secret_not_found	✅
ask	86_configmap_like_but_secret	✅
ask	93_calling_datadog[0]	✅
ask	93_calling_datadog[1]	✅
ask	93_calling_datadog[2]	✅

Legend

• ✅ the test was successful
• :minus: the test was skipped
• ⚠️ the test failed but is known to be flaky or known to fail
• 🚧 the test had a setup failure (not a code regression)
• 🔧 the test failed due to mock data issues (not a code regression)
• 🚫 the test was throttled by API rate limits/overload
• ❌ the test failed and should be fixed before merging the PR