Add model_size experiment to 5 daily workflows; introduce small-agent alias#36997
Merged
Conversation
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
…alias Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Copilot
AI
changed the title
Add
Add Jun 4, 2026
model_size experiment to 5 daily workflows and small-agent model aliasmodel_size experiment to 5 daily workflows; introduce small-agent alias
Copilot created this pull request from a session on behalf of
pelikhan
June 4, 2026 21:00
View session
pelikhan
approved these changes
Jun 4, 2026
Contributor
There was a problem hiding this comment.
Pull request overview
This PR introduces a new built-in model alias (small-agent) and wires a model_size A/B experiment into 5 daily workflow source files to compare small/fast models vs the default agent chain using token-cost and reliability metrics. It also updates several golden fixtures and many generated workflow lockfiles.
Changes:
- Add
small-agentto the built-in model alias map and extend unit tests accordingly. - Add a
model_sizeexperiment + dynamicengine.modelselection to 5 daily workflow source markdown files. - Update WASM golden fixtures and many
*.lock.ymlgenerated workflows (including new Copilot settings injection).
Show a summary per file
| File | Description |
|---|---|
| pkg/workflow/data/model_aliases.json | Adds the small-agent built-in alias chain. |
| pkg/workflow/model_aliases_test.go | Extends built-in alias test coverage to include small-agent. |
| .github/workflows/daily-doc-updater.md | Adds model_size experiment + sets engine.model from activation output. |
| .github/workflows/daily-doc-healer.md | Adds model_size experiment + sets engine.model from activation output. |
| .github/workflows/daily-function-namer.md | Adds model_size experiment + sets engine.model from activation output. |
| .github/workflows/daily-caveman-optimizer.md | Adds model_size experiment + sets engine.model from activation output. |
| .github/workflows/daily-cache-strategy-analyzer.md | Adds model_size experiment + sets engine.model from activation output (codex engine). |
| pkg/workflow/testdata/TestWasmGolden_CompileFixtures/with-imports.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_CompileFixtures/smoke-copilot.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_CompileFixtures/playwright-cli-mode.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_CompileFixtures/basic-copilot.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_AllEngines/pi.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_AllEngines/gemini.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_AllEngines/copilot.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_AllEngines/codex.golden | Updates embedded AWF config golden output to include small-agent. |
| pkg/workflow/testdata/TestWasmGolden_AllEngines/claude.golden | Updates embedded AWF config golden output to include small-agent. |
| .github/workflows/daily-mcp-concurrency-analysis.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-malicious-code-scan.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-issues-report.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-hippo-learn.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-geo-optimizer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-firewall-report.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-file-diet.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-experiment-report.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-compiler-threat-spec-optimizer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-compiler-quality.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-community-attribution.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-cli-tools-tester.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-cli-performance.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-byok-ollama-test.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-assign-issue-to-user.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-architecture-diagram.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-ambient-context-optimizer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/daily-agent-of-the-day-blog-writer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/craft.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/copilot-pr-prompt-analysis.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/copilot-pr-nlp-analysis.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/copilot-pr-merged-report.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/copilot-opt.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/copilot-cli-deep-research.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/contribution-check.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/constraint-solving-potd.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/code-simplifier.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/code-scanning-fixer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/cli-consistency-checker.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/ci-coach.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/chaos-pr-bundle-fuzzer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/breaking-change-checker.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/brave.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/bot-detection.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/auto-triage-issues.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/artifacts-summary.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/architecture-guardian.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/archie.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/agent-persona-explorer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/agent-performance-analyzer.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/ace-editor.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
| .github/workflows/ab-testing-advisor.lock.yml | Injects Copilot settings file write + cleanup trap in run script. |
Copilot's findings
Tip
Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
- Files reviewed: 58/58 changed files
- Comments generated: 3
Comment on lines
+33
to
+40
| engine: | ||
| id: claude | ||
| model: "${{ needs.activation.outputs.model_size }}" | ||
| name: Daily Documentation Updater | ||
| strict: true | ||
| experiments: | ||
| model_size: | ||
| variants: [agent, small-agent] |
Comment on lines
+939
to
+941
| trap 'rm -f /home/runner/.copilot/settings.json' EXIT | ||
| mkdir -p /home/runner/.copilot | ||
| printf '%s' '{"builtInAgents":{"rubberDuck":false}}' > /home/runner/.copilot/settings.json |
Comment on lines
+939
to
+941
| trap 'rm -f /home/runner/.copilot/settings.json' EXIT | ||
| mkdir -p /home/runner/.copilot | ||
| printf '%s' '{"builtInAgents":{"rubberDuck":false}}' > /home/runner/.copilot/settings.json |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Wires a 50/50
model_sizeA/B experiment into 5 eligible daily workflows (no existing experiments, no subagents) to measure whether small models match full-model quality at lower token cost. Adds thesmall-agentbuiltin alias to make the experiment variants resolve at runtime.Model alias
small-agentalias:["haiku", "gpt-5-mini", "gemini-flash", "any"]— parallel toagentbut targeting small/fast modelsExperiment
Added to
daily-doc-updater,daily-function-namer,daily-doc-healer,daily-caveman-optimizer(claude), anddaily-cache-strategy-analyzer(codex):Each workflow's engine is expanded to pass the selected variant as the runtime model: