setup-maf-evals: report generation, three evaluator tiers, MSTest project shape - #2

Closed

leslierichardson95 wants to merge 18 commits into

dev-integrationfrom

skills/setup-maf-evals-v2

Commits on Jun 17, 2026

setup-maf-evals: v2 spec rewrite (MSTest shape, 3 tiers, IChatClient detection)

leslierichardson95
and
Copilot
committed
d44b42f

Commits on Jun 18, 2026

setup-maf-evals: dogfood findings from ELI5Agent run

leslierichardson95
and
Copilot
committed
6a166c7
setup-maf-evals: per-run metrics glossary + friendly user-secrets diagnostic

leslierichardson95
and
Copilot
committed
8255ef8
setup-maf-evals: slim SKILL.md by moving prose into references/

leslierichardson95
and
Copilot
committed
012a9a6
setup-maf-evals: factory must explicitly load user-secrets in test hosts

leslierichardson95
and
Copilot
committed
cc57ef0
setup-maf-evals: warn about reasoning models rejecting max_tokens

leslierichardson95
and
Copilot
committed
b7ca903

Commits on Jun 22, 2026

setup-maf-evals: quality-first chat output + rubric-driven evaluator + compare opt-in

leslierichardson95
and
Copilot
committed
4695899
setup-maf-evals: fix RubricEvaluator template raw-string interpolation (CS9006)

leslierichardson95
and
Copilot
committed
40606ef
docs(setup-maf-evals): explain Cache Miss in Diagnostic Data

leslierichardson95
and
Copilot
committed
dbd46ef
feat(setup-maf-evals): route agent call through cached client

leslierichardson95
and
Copilot
committed
2959158
docs(dotnet-ai): add common-pitfalls + check-id glossary across perf skills

leslierichardson95
and
Copilot
committed
ee31991
docs(scan-agentic-app-perf): escape placeholder tokens in checks references

leslierichardson95
and
Copilot
committed
6769639
feat(select-agent-models): add greenfield 'plan' sub-mode

leslierichardson95
and
Copilot
committed
5dffeff

Commits on Jun 24, 2026