agentleFS
Sign inStart free
e-dule-dul/opencl-lab/.claude/skills/grading/SKILL.md

grading

Skill0 starsChanged 8 months agoMITView on GitHub

Standardized grading rubric for evaluating OpenCL Lab modules.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Big picture: Read workflow/design/00-executive-summary.md to understand vision, philosophy and structure.

    line 9
  2. Very specific:

    Analyze Content: Read the module's *README*.md and scan the relevant .cpp, .h, .cl, and CMakeLists.txt files.

    line 10
  3. Very specific:

    When directory does't have dedicated .md check parent directory.

    line 11
  4. Distinctive:

    Code-to-Explanation Clarity: Do the code comments and README explain the *WHY* behind the architectural choices (RAII, cl.hpp) rather than …

    line 19

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Course Grading Skill

Instructions

  1. Big picture: Read workflow/design/00-executive-summary.md to understand vision, philosophy and structure. Then read .claude/rules/00_master_specs.md to understand design decisions and known limitations.
  2. Analyze Content: Read the module's *README*.md and scan the relevant .cpp, .h, .cl, and CMakeLists.txt files.
  3. When directory does't have dedicated .md check parent directory.
  4. Apply Rubric: Grade the module on a scale of 1 to 10 for each of the following criteria. 1 is terrible, 10 is perfect.

Grading Criteria

  • Theory to Application Ratio: Does the module strike the right balance? (e.g., Is theory introduced "Just-in-Time" to solve a real application problem, or is it a dry theory dump?)
  • Uniqueness: Does this module offer unique value (e.g., custom ISP, ROS2 integration) compared to standard, generic OpenCL tutorials found online?
  • Hardware Dependency: Is the code portable? (e.g., Does it gracefully handle OpenCL 1.2 fallbacks, or does it strictly require an Nvidia GPU with OpenCL 2.0 SVM?)
  • Repetitiveness: Is the material concise, or does the README redundantly state things that are already obvious from the code comments?
  • Code-to-Explanation Clarity: Do the code comments and README explain the WHY behind the architectural choices (RAII, cl.hpp) rather than just stating WHAT the code does?
  • Reproducibility: Is the build process (CMake) entirely self-contained and easy for a student to execute without external environment hell?

Final Calculation

  • Calculate the average of all criteria to provide a Summarized Score (1-10) at the bottom of the report.

Strict Output Template

Output exactly in this Markdown table format:

Module NameTheory/AppUniquenessHW Dep.RepetitivenessClarityReproducibilityFINAL SCORE
[Module 1][x]/10[x]/10[x]/10[x]/10[x]/10[x]/10[Avg]/10

Justification

For each criterion, write the Score and a 1-2 sentence short Justification:

  • Theory/App (8) — BT.601 math and T-API ordering are placed just-in-time, but Key Concepts section lacks forward references to exact code locations.
  • Uniqueness (9) — NV12/YUYV decoding, RemoteTenso
  • ....

Actionable Items

Actionable items to addres potential issues

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free