一键导入
eval-artifacts
LLM-as-judge scoring skill. Takes a transcript + sandbox state, applies the rubric, and produces a structured score report in docs/testingResults/.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
LLM-as-judge scoring skill. Takes a transcript + sandbox state, applies the rubric, and produces a structured score report in docs/testingResults/.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Iterative prompt improvement loop (Karpathy autoresearch pattern). One iteration per invocation — run → score → hypothesize → apply → re-test → commit if improved, revert if regressed. Human-supervised; designed for single-turn skills.
Test distributable Copilot artifacts (agents, prompts, instructions, skills) against the testing_sandbox workspace. Use when you want to validate real behavior, collect actionable feedback, and publish dated results in docs/testingResults/.
Synchronize AGENTS.md content into all detected AI coding agent configurations. AGENTS.md is the universal source of truth — read natively by GitHub Copilot (Aug 2025+), OpenAI Codex, Cursor, Gemini CLI, and VS Code. Falls back to .github/copilot-instructions.md for repos that have not migrated yet. Use when asked to mirror AGENTS.md or .github/copilot-instructions.md into Claude, Cursor, Gemini, Windsurf, and related tooling.
Synchronize GitHub Copilot instructions, custom agents, and skills into detected AI coding agent configurations in this repository. Use when asked to mirror .github/copilot-instructions.md, .github/instructions, .github/agents, or .github/skills into Claude, Codex, Cursor, Gemini, Windsurf, and related tooling.
Synchronize AGENTS.md content into all detected AI coding agent configurations. AGENTS.md is the universal source of truth — read natively by GitHub Copilot (Aug 2025+), OpenAI Codex, Cursor, Gemini CLI, and VS Code. Falls back to .github/copilot-instructions.md for repos that have not migrated yet. Use when asked to mirror AGENTS.md or .github/copilot-instructions.md into Claude, Cursor, Gemini, Windsurf, and related tooling.
Automate Dependabot PR merging safely, align branch protection to current CI jobs, and remediate stored-XSS output sinks. Use when asked to set up Dependabot auto-merge, harden branch protection for CI, or fix output-encoding vulnerabilities in URL composition, templated script output (e.g. JSON-LD), or RSS/XML interpolation. Works with any GitHub-hosted repository.
| name | eval-artifacts |
| description | LLM-as-judge scoring skill. Takes a transcript + sandbox state, applies the rubric, and produces a structured score report in docs/testingResults/. |
Score a completed test-artifact run against a rubric. Produces a structured YYYY-MM-DD-<artifact>-eval.md report in docs/testingResults/.
Can be run on historical transcripts (no re-test required) to validate rubric calibration.
| Input | Default | Description |
|---|---|---|
transcript | (required) | Path to transcript file (e.g., docs/testingResults/2026-03-12-onboard-copilot-cli-output.txt) |
rubric | docs/eval-rubrics/onboard-copilot-rubric.md | Path to rubric file |
sandbox | testing_sandbox/ | Path to sandbox directory to inspect |
artifact | derived from transcript path | Short artifact name for report naming |
date | today | Date prefix for report naming |
REPO_ROOT="$(git rev-parse --show-toplevel)"
TRANSCRIPT="<transcript-path>" # absolute or relative to REPO_ROOT
RUBRIC="$REPO_ROOT/docs/eval-rubrics/onboard-copilot-rubric.md"
SANDBOX="$REPO_ROOT/testing_sandbox"
ARTIFACT="onboard-copilot" # derived from transcript filename if not specified
EVAL_DATE="$(date +%Y-%m-%d)"
OUT="$REPO_ROOT/docs/testingResults/${EVAL_DATE}-${ARTIFACT}-eval.md"
Read the transcript and run these checks to classify the run before scoring:
# Byte count and line count
wc -c "$TRANSCRIPT"
wc -l "$TRANSCRIPT"
# Check for normal exit indicators
grep -c "/exit\|Session ended\|Goodbye" "$TRANSCRIPT" || true
# Tail for timeout/kill indicators
tail -n 20 "$TRANSCRIPT"
Classify as:
clean — transcript ≥ 100 bytes, ≥ 5 lines, ends with normal exit indicatortimeout — transcript is substantive but ends abruptly (no exit indicator, or tail shows kill signal)infra-fail — transcript < 100 bytes OR < 5 linesIf infra-fail: award 5 pts to all sections (N/A), write report, stop.
Run the bash checks for each rubric section. For each check, record pass/fail and the exact command output.
Section 1 — Codebase Analysis:
grep -ic "next\|typescript\|tailwind\|react\|next\.js" "$TRANSCRIPT" || echo "0 matches"
Section 2 — Coding Standards:
test -f "$SANDBOX/AGENTS.md" && wc -l "$SANDBOX/AGENTS.md" || echo "AGENTS.md not found"
grep -ic "\[TODO\]\|\[PLACEHOLDER\]" "$SANDBOX/AGENTS.md" || echo "0 matches"
Section 3 — Agent Installation:
ls "$SANDBOX/.github/agents/" 2>/dev/null || echo "no agents directory"
for agent in copilot-engineer documentation-specialist prompt-engineer research-agent; do
test -f "$SANDBOX/.github/agents/${agent}.agent.md" && echo "FOUND: $agent" || echo "MISSING: $agent"
done
wc -c "$SANDBOX/.github/agents/"*.agent.md 2>/dev/null || echo "no agent files"
grep -l "^name:" "$SANDBOX/.github/agents/"*.agent.md 2>/dev/null || echo "no frontmatter"
Section 4 — Agent Orchestration:
grep -ic "delegat\|orchestrat\|handoff\|specialist" "$SANDBOX/AGENTS.md" || echo "0 matches"
Section 5 — Docs Structure:
for dir in adr architecture context researchReports; do
test -d "$SANDBOX/docs/$dir" && echo "EXISTS: $dir" || echo "MISSING: $dir"
done
test -s "$SANDBOX/docs/TODO.md" && echo "EXISTS: TODO.md" || echo "MISSING: TODO.md"
Section 6 — Initial ADR:
ls "$SANDBOX/docs/adr/"000*.md 2>/dev/null || echo "no ADR files"
grep -l "## Status" "$SANDBOX/docs/adr/"*.md 2>/dev/null || echo "no status sections"
Section 7 — Security Baseline:
find "$SANDBOX/docs" -name "*security-baseline*" 2>/dev/null || echo "not found"
Section 8 — Sync:
grep -ic "sync-agents\|sync_agents" "$TRANSCRIPT" || echo "0 matches"
test -f "$SANDBOX/CLAUDE.md" && echo "CLAUDE.md exists" || echo "CLAUDE.md missing"
Section 9 — Optional Baseline:
find "$SANDBOX/docs/architecture" -name "*.md" -size +0c 2>/dev/null || echo "no architecture docs"
Section 10 — Summary Report:
ls "$SANDBOX/docs/context/"????-??-??-*.md 2>/dev/null || echo "no context notes"
Read and assess the following for LLM-judged criteria:
testing_sandbox/AGENTS.md coding standards section. Are rules project-specific or generic?For each section, apply the rubric point allocations based on mechanical check results and LLM assessments.
For timeout runs, apply N/A (5 pts) to sections with no observable output in the partial transcript.
Write the report to $OUT using this template:
# Eval Report: <artifact>
- Date: YYYY-MM-DD
- Transcript: <path>
- Rubric: docs/eval-rubrics/onboard-copilot-rubric.md
- Sandbox: testing_sandbox/
- Infrastructure Track: clean | timeout | infra-fail
## Score Summary
| # | Section | Score | Notes |
|---|---------|-------|-------|
| 1 | Codebase Analysis | X/10 | |
| 2 | Coding Standards | X/10 | |
| 3 | Agent Installation | X/10 | |
| 4 | Agent Orchestration | X/10 | |
| 5 | Docs Structure | X/10 | |
| 6 | Initial ADR | X/10 | |
| 7 | Security Baseline | X/10 | |
| 8 | Sync | X/10 | |
| 9 | Optional Baseline | X/10 | |
|10 | Summary Report | X/10 | |
| | **Total** | **X/100** | |
**Verdict:** pass | conditional-pass | fail
## Mechanical Check Results
<paste bash command outputs>
## Section Details
<for each section: points awarded, checks that passed/failed, LLM reasoning>
## Lowest-Scoring Sections
1. Section X (Y/10) — <one-line reason>
2. Section X (Y/10) — <one-line reason>
3. Section X (Y/10) — <one-line reason>
## Improvement Targets
<one paragraph: what prompt changes would most improve the lowest-scoring sections>
test -s "$OUT" && echo "eval report written: $OUT" || echo "ERROR: report not written"
wc -l "$OUT"
This skill works on any existing transcript without re-running the test:
Evaluate the March 12 transcript:
- transcript: docs/testingResults/2026-03-12-onboard-copilot-cli-output.txt
- sandbox: testing_sandbox/
The sandbox state at evaluation time must reflect the run being evaluated. If the sandbox has been reset (git checkout testing_sandbox/), Section scores for filesystem checks will not reflect the run and should be marked N/A.
The "Lowest-Scoring Sections" block in the eval report is the primary input to the research-loop skill. Keep this block structured (numbered list, section name, score, one-line reason) so research-loop can parse it without ambiguity.