跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
演示文稿

submission-auditor

Specification-driven submission auditing for university professors. 4-agent team that compiles a professor's template, requirements doc, rubric, or exemplar into a checkable Submission Spec, audits student submissions (single or batch) against it with evidence-located findings, and produces per-student feedback reports plus a class-level pattern report. Works for any genre: lab reports, papers, theses, code projects, presentation decks. Triggers on: check submissions, format check, check against template, does this meet the requirements, lab report check, audit student work, compliance check, submission requirements, batch check, 检查作业, 格式检查, 格式审查, 实验报告检查, 论文格式, 检查是否符合要求, 批量检查, 提交规范.

YujxZJCN · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add YujxZJCN/teaching-skills --skill submission-auditor

技能说明


name: submission-auditor description: "Specification-driven submission auditing for university professors. 4-agent team that compiles a professor's template, requirements doc, rubric, or exemplar into a checkable Submission Spec, audits student submissions (single or batch) against it with evidence-located findings, and produces per-student feedback reports plus a class-level pattern report. Works for any genre: lab reports, papers, theses, code projects, presentation decks. Triggers on: check submissions, format check, check against template, does this meet the requirements, lab report check, audit student work, compliance check, submission requirements, batch check, 检查作业, 格式检查, 格式审查, 实验报告检查, 论文格式, 检查是否符合要求, 批量检查, 提交规范." metadata: version: "1.0.0" last_updated: "2026-06-10" status: active pipeline_stage: 4 related_skills: - assessment-architect - student-mentor - teaching-pipeline

Submission Auditor — Specification-Driven Submission Review

Turns "here's my standard, check their work against it" into a disciplined three-step flow: compile the spec → audit with located evidence → report at two altitudes (per-student feedback, class-level patterns). The professor's standard is the law; this skill makes it checkable, applies it consistently across every submission, and reports honestly about which checks were mechanical and which were judgment.

Prime rule: an audit finding without a location in the student's work does not exist. Every finding cites where it looked ("§3 Results, p.2: figure has no caption"), and every judgment-type finding quotes the evidence it judged. "Generally weak formatting" is not a finding.

Quick Start

Here's my lab report template — check these 40 submissions against it
这是我的实验报告模板,帮我检查这批学生提交的格式和内容是否符合要求
Does this thesis draft meet our department's format requirements? (attach both)
Build a submission checklist students can self-check before the deadline

Modes

ModeTrigger intentOutput
spec"Here's my template/requirements" (first contact with a new standard)Compiled Submission Spec — every requirement made checkable, classified deterministic vs judgment, confirmed at a checkpoint
audit"Check this submission" with a spec (or material to compile one)Audit findings + per-student feedback report draft
batch-audit"Check these N submissions"Per-student reports + class-level pattern report + consistency record
self-check"Make a checklist students can use themselves"Student-facing self-check version of the spec (deterministic checks only, supportive phrasing)
calibrate"The audit was too strict/lenient on X" or professor reviews a sampleSpec revision pass: tighten/loosen checks against professor verdicts on sampled findings; revisions logged in the spec

Mode dispatch rule: no confirmed spec yet → spec mode always runs first, whatever was asked. Auditing against an unconfirmed standard produces authoritative-looking reports built on guesses — the blueprint-first rule of assessment-architect, applied to checking.

Does NOT trigger

ScenarioUse instead
Designing the assignment, template, or rubric itselfassessment-architect
Qualitative feedback beyond the spec (style, argument coaching)student-mentor feedback
Grading / assigning scoresThis skill audits compliance; scoring is the professor's act (see Iron Rule 4)
Checking AI-resilience of the assignment designassessment-architect integrity-check

Agent Team (4)

AgentRole
spec_compiler_agentCompiles template/requirements/rubric/exemplar into the Submission Spec; classifies checks deterministic vs judgment; pushes ambiguous requirements back to the professor instead of guessing
format_auditor_agentRuns deterministic checks: structure, required sections, length, citation format presence, figures/tables/captions, naming, file conventions — location-cited PASS/FAIL/NOT_EVALUABLE per check
content_auditor_agentRuns judgment checks: content requirements judged against quoted evidence from the submission; every verdict labeled JUDGMENT with confidence; never drifts beyond the spec
report_writer_agentTwo-altitude reporting: per-student feedback report (formative, per student-mentor/references/feedback_principles.md structure) + class-level pattern report; runs the batch fairness pass

Workflow (batch-audit mode)

Phase 0  SPEC       — load confirmed Submission Spec (or run `spec` mode first)
         🧑 checkpoint: spec confirmed — which checks are deterministic, which are
            judgment, severities, and what the professor chose NOT to check
Phase 1  AUDIT      — per submission: format_auditor (deterministic) then
                      content_auditor (judgment) → findings ledger per student,
                      every finding location-cited
Phase 2  FAIRNESS   — report_writer cross-submission pass: same defect ⇒ same
                      severity and same wording class everywhere; judgment-check
                      verdicts spot-compared across submissions for drift
Phase 3  REPORT     — per-student feedback reports (drafts) + class pattern report
                      (defect frequencies, common misses, spec ambiguities surfaced
                      by the data)
         🧑 checkpoint: professor reviews JUDGMENT findings sample + pattern report
            before any per-student report is released

The class pattern report routes two ways: high-frequency misses are a teaching signal (offer a lesson-builder remediation segment or an announcement draft), and recurring spec ambiguities are a spec signal (offer calibrate mode). Both feed passport iteration_history as evidence — with no student-identifying data (counts only).

Iron Rules

  1. Spec before audit. No submission is audited
阅读完整来源 ↗