[skills-eval] dotnet-ai: 5 skills, 12% pass — 1 P1, 3 to strengthen, 1 keep

[AbhitejJohn]

Context — cross-family skill evaluation: dotnet-ai

This issue is self-contained: it captures everything a skill author needs to act on the dotnet-ai plugin without opening the full report.

What this measures. Every runnable skill in dotnet/skills was run through Vally (0.7) on a cross-family matrix: 5 executor model families — opus-4.8, gpt-5.5, sonnet-4.6, haiku-4.5, mai-flash — each judged by a different family (judge ≠ executor; default judge = latest Opus, or GPT when Opus is the executor). For every skill, a skilled run is compared against a baseline (no-skill) run and scored per executor. This removes single-model and self-judging bias, so a skill that only helps one model family — or only its own family's judge — is visible.

• Data source: cross-family CI grid (run 29228914412 + backfills) — 5 executors × 85 runnable skills, 419 scored cells over 84 skills.
• Row grain: one row per skill, aggregated across its (up to 5) executor cells. avgN is the mean trial count behind the cells (trials 1–17; higher = more statistically trustworthy). thin-N flags directional-only rows.

dotnet-ai at a glance (portfolio scorecard)

Plugin	Skills	Cells	Pass	Impact	Tie-trials	Err	avg ΔTok	avg ΔTurns	avg ΔTools	Headline
dotnet-ai	5	25	12%	0.489	10	0	+16,639	−0.14	−0.48	Underrated: high impact, efficient, low pass

Insights. 5 skill(s); mean impact 0.49; 1/5 help ≥1 frontier model. No frontier-validated exemplar. Weak-model-only wins (frontier misses): mcp-csharp-create.

Address first: technology-selection — cost.

How to read the table

Each skill is scored skilled vs. baseline on these axes:

Signal	Column	What it means	Good
Breadth	Families ✓ (n/5)	How many of the 5 model families the skill helps (per-family pass count, max 5/5)	4–5 / 5
Where	Passed on	Which families passed. Frontier = latest Opus + latest GPT are bold; Sonnet 4.6 / Haiku 4.5 / MAI Flash are mid/low-weight	frontier ✓
Magnitude	Impact (−1…+1)	How strongly the judge prefers skilled over baseline	≥ 0.4
Decisiveness	Ties▵	Trials where the judge saw no difference → skill is inert	low
Safety	Loss▵	Trials where skilled was WORSE than baseline → skill misfires	~0
Reliability	Err	Trials that errored/crashed in setup or judging	0
Efficiency	ΔTok / ΔTurns / ΔTools	Extra tokens / agent turns / tool calls vs baseline	≤ 0
Confidence	avgN	Mean trials behind the verdict; low N = directional only	≥ 3
Invocation	Call%	Share of skilled trials where the model actually invoked the skill	~100%

Families ✓ is cell-level (max 5/5); Ties▵/Loss▵ are trial-level tallies summed across all families (including the ones where the skill failed). A high Families ✓ next to non-zero Loss▵ is not a contradiction — see Passed on and the Action text for where losses landed.

Action buckets (each skill has one primary action; [flags] note secondary concerns):

Bucket	Priority	Meaning
FIX-RELIABILITY	🔴 P0	Errored trials / no verdict — stabilize the harness before trusting the score
FIX-DISCOVERY	🔴 P0	Model doesn't invoke it (Call% < 50%) — a triggering/description problem
FIX-REGRESSION	🔴 P0	Skilled is worse than baseline on many trials — the skill misfires
ADD-DECISIVENESS	🟠 P1	Called ~100% but ties dominate, ~0 impact — inert; needs sharper behavioral steps
TRIM-COST	🟠 P1	Passes but with heavy token/turn overhead — trim verbosity
EXEMPLAR	🟢 keep	Broad, strong, reliable win — use as a template
EFFICIENT-WIN	🟢 protect	Wins and cuts turns/tools — the ideal shape
KEEP-POLISH	🟢	Solid majority win; minor polish + more trials
STRENGTHEN	🟡 P2	Marginal/mixed lift — sharpen triggers & success criteria

Per-skill actions

Skill	Families ✓	Passed on (frontier bold)	Impact	Ties▵	Loss▵	Err	avgN	Call%	ΔTok	ΔTurns	ΔTools	Action
mcp-csharp-test	0/5	—	0.65	0	1	0	2.8	87%	-36,898	-0.73	-0.87	STRENGTHEN · 🟡 Marginal/mixed (passes 0/5). Diagnosis: misses both frontier models — likely assumes context they solve unaided. Try: (1) sharpen the trigger so it fires only where it wins; (2) add 1–2 opinionated, concrete steps that change behaviour; (3) add trials to separate signal from noise. If frontier models never benefit, scope it explicitly to weaker models or reconsider its value. ([both-frontier-miss])
technology-selection	1/5	GPT	0.57	1	0	0	3.6	100%	+208,946	+3.73	+3.47	TRIM-COST · 🟠 Passes but heavy (+208,946 tok). Value is real; cost isn't justified. Trim: (1) cut redundant/whole-file reads — point to specific sections; (2) replace "explore everything" with a targeted checklist; (3) move deep reference material behind links instead of inlining it. ([frontier-miss], [costly])
mcp-csharp-publish	0/5	—	0.56	1	1	0	3	100%	-18,308	-0.93	-1.47	STRENGTHEN · 🟡 Marginal/mixed (passes 0/5). Diagnosis: mostly ties — too generic/non-prescriptive. Try: (1) sharpen the trigger so it fires only where it wins; (2) add 1–2 opinionated, concrete steps that change behaviour; (3) add trials to separate signal from noise. If frontier models never benefit, scope it explicitly to weaker models or reconsider its value. ([both-frontier-miss])
mcp-csharp-create	2/5	Sonnet, Haiku	0.40	2	2	0	2.6	100%	-94,279	-4.3	-5.53	EFFICIENT-WIN · ✅ Better and cheaper. Wins on Sonnet, Haiku while saving 4.3 turns, 5.53 tools, 94,279 tok. What's good: it adds value without bloat — the ideal shape. Protect the brevity; don't let it grow. ⚠️ But both frontier models miss and regress — this may only be helping mid/low-weight models. Treat as a weak-model win, not a universal one; check the frontier trajectories before templating. ([both-frontier-miss], [frontier-regressed], [efficient])
mcp-csharp-debug	0/5	—	0.27	6	1	0	3	100%	+23,734	+1.53	+2	STRENGTHEN · 🟡 Marginal/mixed (passes 0/5). Diagnosis: mostly ties — too generic/non-prescriptive. Try: (1) sharpen the trigger so it fires only where it wins; (2) add 1–2 opinionated, concrete steps that change behaviour; (3) add trials to separate signal from noise. If frontier models never benefit, scope it explicitly to weaker models or reconsider its value. ([both-frontier-miss], [frontier-regressed])

---

_{Generated from the cross-family Call-to-Action report (CALL-TO-ACTION.md §4–§5; companion IMPACT-ANALYSIS.md). Regenerate the underlying tables with node deep-metrics.mjs "$env:TEMP\cf-ci" agg-ci → node gen-cta-tables.mjs agg-ci. Numbers are directional where avgN is low; treat single-trial cells as hypotheses to confirm with more runs.}

[github-actions]

🎯 Issue Triage

Summary: This is a cross-family skill evaluation report for the dotnet-ai plugin covering 5 skills (mcp-csharp-test, mcp-csharp-publish, mcp-csharp-debug, mcp-csharp-create, technology-selection). The portfolio has 12% pass rate with one P1 cost issue (technology-selection), three skills needing strengthening, and one efficient-win to protect (mcp-csharp-create).

Type: task (evaluation work item with actionable per-skill improvements)

Assigned owners: @leslierichardson95 and @cathysull (mcp-csharp-* skills), @luisquintanilla and @artl93 (technology-selection)

Confidence: High — clear area match to dotnet-ai, owners unambiguous from CODEOWNERS.

Similar issues: No other open [skills-eval] dotnet-ai issues found.

Triage details

• Area mapping: All skills are in /plugins/dotnet-ai/, mapped to area-ai label (repository uses area-ai; area-dotnet-ai does not exist).
• CODEOWNERS match: mcp-csharp-* → @leslierichardson95 @cathysull; technology-selection → `@luisquintanilla `@artl93
• Confidence rationale: Issue is auto-generated from CI eval run; area, type, and owners are all unambiguous.
• Suggested next steps: Prioritize technology-selection TRIM-COST (P1), then strengthen the three 0/5 skills. mcp-csharp-create should be protected and not expanded.

Generated by Issue Triage for issue #889 · 91.9 AIC · ⊞ 28.5K · ◷

[AbhitejJohn]

Verifying fixes for this plugin's skills

To re-run the cross-family evaluation and confirm a change moves the pass rate, cherry-pick the dispatch workflow from #920 and dispatch it scoped to this plugin:

git fetch origin abhitejjohn-cross-family-eval-workflow
git cherry-pick 7c97b7346          # single commit in #920
git push origin HEAD

gh workflow run evaluation-run.yml --ref <your-branch> \
  -f plugins="dotnet-ai" \
  -f executors="opus,gpt,sonnet46,haiku,mai" -f runs=3

To verify a single skill faster, scope with skills (the skill directory name, e.g. one of the skills listed above):

gh workflow run evaluation-run.yml --ref <your-branch> \
  -f skills="<skill-dir>" -f executors="opus,gpt" -f runs=3

Each skill is run skilled-vs-baseline across the 5-family executor matrix, with the judge never in the executor's family (default judge = latest Opus; GPT when Opus executes). Results assemble into a cross-family pass matrix you can compare against the numbers in this issue. Full inputs/usage: #920.