[skills-eval] dotnet-diag: 7 skills, 49% pass — 1 P0, 1 P1, 3 to strengthen, 2 keep

[AbhitejJohn]

🔄 EDITED 2026-07-19 — correction per #909

This finding has been revised. The errored trials that drove the FIX-RELIABILITY (P0) flag(s) below were shown in #909 to be judge-side infrastructure failures — a disabled/throttled PAT (CAPIError 400 organization disabled) and Vally's session.idle judge timeout — not fixture nondeterminism. Verdict scores already exclude errored trials, so re-running the classifier with those judge errors dropped reclassifies the affected skill(s):

Skill	Was	Now
analyzing-dotnet-performance	FIX-RELIABILITY (P0)	KEEP-POLISH (keep)
android-tombstone-symbolication	FIX-RELIABILITY (P0)	FIX-REGRESSION (P0, real)

Struck-through text below is the superseded finding — in particular the "pin fixture SDK/tool versions" remediation does not apply. The title's P0 count has been updated. No eval re-run is required to correct the scores; recovering the lost judgments (optional) means re-judging those trials, not re-running the skill.

Context — cross-family skill evaluation: dotnet-diag

This issue is self-contained: it captures everything a skill author needs to act on the dotnet-diag 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-diag at a glance (portfolio scorecard)

Plugin	Skills	Cells	Pass	Impact	Tie-trials	Err	avg ΔTok	avg ΔTurns	avg ΔTools	Headline
dotnet-diag	7	35	49%	0.362	33	2	+35,515	+0.74	+2.50	Half-inert (33 ties); regressions in symbolication

Insights. 7 skill(s); mean impact 0.36; 4/7 help ≥1 frontier model. Exemplars (win incl. frontier): dotnet-trace-collect. Weakest: android-tombstone-symbolication (0.16).

Address first: analyzing-dotnet-performance — reliability; android-tombstone-symbolication — reliability; microbenchmarking — cost.

Address first: android-tombstone-symbolication — regression (reclassified from reliability); microbenchmarking — cost. (analyzing-dotnet-performance reclassified out of P0 — see EDIT above.)

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
dotnet-trace-collect	4/5	Opus, Sonnet, Haiku, MAI	0.53	8	8	0	17	97%	+30,164	+1.48	+1.6	EXEMPLAR · ✅ Template-worthy. What's good: broad (passes 4/5, incl. frontier Opus). Keep as-is; lift its structure (crisp triggers + imperative steps) into weaker siblings. ⚠️ Caveat: frontier miss on GPT — confirm it isn't only lifting weaker models on cases frontier already handles. ([frontier-miss])
clr-activation-debugging	2/5	Opus, MAI	0.44	2	5	0	6.8	94%	+39,418	+1.09	+2.36	STRENGTHEN · 🟡 Marginal/mixed (passes 2/5 — Opus, MAI). Diagnosis: model-dependent — helps some families, not others. 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. ([frontier-miss])
microbenchmarking	4/5	Opus, GPT, Sonnet, MAI	0.44	1	0	0	1	100%	+237,705	+5.6	+9.2	TRIM-COST · 🟠 Passes but heavy (+237,705 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. ([costly], [thin-N])
dump-collect	2/5	Sonnet, Haiku	0.37	8	8	0	9	80%	+18,779	+1.11	+1.6	STRENGTHEN · 🟡 Marginal/mixed (passes 2/5 — Sonnet, Haiku). 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])
analyzing-dotnet-performance	3/5	GPT, Haiku, MAI	0.33	7	4	1	9.6	96%	+5,866	-0.73	+6.33	FIX-RELIABILITY · 🔴 1 errored trial(s) — verdict can't be trusted until setup is deterministic. This is an infra/harness fix, not a content one: (1) capture the failing trial's stderr; (2) pin tool/SDK versions in the fixture; (3) add a setup smoke-check before scoring. ([reliability], [frontier-miss]) 🔄 EDIT (see #909): KEEP-POLISH · 🟢 Solid majority win (3/5 — GPT, Haiku, MAI). The 1 errored trial was a judge-side session.idle timeout (sonnet46), not fixture flakiness. Investigate the Opus miss; minor polish. ([frontier-miss])
apple-crash-symbolication	1/5	Sonnet	0.26	5	1	0	2.8	67%	-75,504	-3.33	-3.37	STRENGTHEN · 🟡 Marginal/mixed (passes 1/5 — Sonnet). 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])
android-tombstone-symbolication	1/5	Haiku	0.16	2	13	1	6.6	88%	-7,821	-0.08	-0.2	FIX-RELIABILITY · 🔴 1 errored trial(s) — verdict can't be trusted until setup is deterministic. This is an infra/harness fix, not a content one: (1) capture the failing trial's stderr; (2) pin tool/SDK versions in the fixture; (3) add a setup smoke-check before scoring. ([reliability], [both-frontier-miss], [frontier-regressed]) 🔄 EDIT (see #909): FIX-REGRESSION · 🔴 Still P0 — but the real problem is regression, not reliability. The 1 errored trial was a judge-side disabled-PAT failure in the mai cell; removing it unmasks losses on both frontier models (Opus, GPT) — cells that had no errored trials (aggregate loss ~39%). Add explicit stop-conditions and narrow the trigger. ([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-diag plugin (7 skills, 49% pass rate, mean impact 0.36). It identifies 2 P0 reliability issues (analyzing-dotnet-performance, android-tombstone-symbolication), 1 P1 cost issue (microbenchmarking), 3 skills to strengthen, and 1 exemplar (dotnet-trace-collect). The area is dotnet-diag (diagnostics, performance, debugging).

Type: task — a skill improvement work item generated from evaluation data.

Assigned owners: @webreidi and @AbhitejJohn assigned directly. The teams dotnet/dotnet-diag and dotnet/skills-diag-reviewers are the primary CODEOWNERS but cannot be set as direct assignees.

Confidence: High — clear area mapping, obvious task type, unambiguous CODEOWNERS match.

Similar issues: No duplicate open issues found for this evaluation cycle.

Triage details

• Area mapping: plugins/dotnet-diag/ maps to area-dotnet-diag; the existing area-diag label was also retained.
• CODEOWNERS match: /plugins/dotnet-diag/ → dotnet/dotnet-diag + dotnet/skills-diag-reviewers; specific skills have additional owners (brianrob/noahfalk for dotnet-trace-collect, marklio/ChrisAhna for clr-activation-debugging, dotnet/dotnet-perf-team for microbenchmarking).
• Confidence rationale: Machine-generated eval report with well-defined action buckets; type is clearly task.
• Suggested next steps: P0 items (analyzing-dotnet-performance, android-tombstone-symbolication) should be addressed first to stabilize harness reliability before trusting scores. microbenchmarking TRIM-COST is P1.

Generated by Issue Triage for issue #893 · 115.1 AIC · ⊞ 28.4K · ◷

[leculver]

I do not own this, you should find the appropriate owner.

[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-diag" \
  -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.