design-compare
Skill19 starsChanged 21 days agoApache-2.0View on GitHub
Use when the user wants to compare two or more lens designs side by side. Triggers on "compare", "which is better", "side by side", "A vs B", "rank these designs".
3 other projects use a version of this.
Adopted by
Most recently taken up by
- Aboudjem/ui-ux-suitetheir own version, first seen Nov 8, 2026
- artemnovichkov/skillstheir own version, first seen Aug 6, 2026
- DaniAkash/skillstheir own version, first seen May 19, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- artemnovichkov/skills★ 42same name
- Aboudjem/ui-ux-suite★ 8same name
- DaniAkash/skills★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- AGENTS.mdAGENTS.md
- bench-catalogSkill
- design-aberration-reviewSkill
- design-glass-matchingSkill
- design-reviewSkill
- design-tolerance-desensitizeSkill
- design-vision-reviewSkill
- optimize-loopSkill
- zos-api-debugSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 6 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 17
After running optimization from different starting points
- Very specific: line 19
NOT for: single-design assessment (use /design-review)
- Distinctive: line 37
Read system info — EFL, F-number, total track, element count.
- Distinctive: line 89
After the comparison, reload whichever design the user was working on before it started.
- Distinctive: line 96
Same analysis on every design: Don't skip MTF on one and run it on another.
- Distinctive: line 101
Never drop a field from the comparison because it is bad: Every field you analyzed appears in the report, worst included.
- Distinctive: line 108
If comparing more than 4 designs, suggest narrowing down first — too many columns gets unreadable.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.
Files
Design Compare — Side-by-Side Lens Comparison
Overview
Loads multiple designs, runs an identical analysis on each through the ZOS-API, and outputs a comparison table highlighting which design wins on each metric. The user makes the final call — this skill organizes the data and recommends.
When to Use
- User has 2+ designs and wants to pick the best one
- After running optimization from different starting points
- Comparing a design before/after optimization
- NOT for: single-design assessment (use
/design-review)
Workflow
Step 1: Identify designs to compare
Get the list from the user. Formats:
- File paths to
.zmxfiles (e.g.data/lenses/design_a.zmx,design_b.zmx). - A sample / built-in design from the OpticStudio sample library.
- Current design vs a saved file ("compare current state with the saved version").
Step 2: Analyze each design
For each design in sequence:
- Load it (or skip if "current").
- Read system info — EFL, F-number, total track, element count.
- Pull RMS spot for all fields.
- Pull MTF at a representative frequency for all fields.
- Record the results, move to the next design.
Decide the battery before you run the first design, and run that same battery on every design. Adding an analysis after you have moved on costs a reload and invites an asymmetric comparison.
get_mtf returns one more data series than you have fields. The FIRST series
(index 0) is the diffraction limit, not a field, when the result reports
diffraction_limit_shown true — it is the same curve for both designs and comparing it
tells you nothing. When diffraction_limit_shown is false, series 0 is the first field;
when it is null, do not map the series by position at all. Count the series against the field
count before you map them: with 3 fields you get 4 series, with 4 fields you get 5.
On-axis is series 1, and the last series is the corner. Two designs with
different field counts have different series counts — map each one separately.
Important: save the current design FIRST if it has unsaved changes, so it can be reloaded after the comparison.
Step 3: Output comparison table
## Design Comparison
| Metric | Design A | Design B | ... | Winner |
|----------------|----------|----------|-----|--------|
| EFL | ... | ... | ... | — |
| F-number | ... | ... | ... | — |
| Total track | ... | ... | ... | shorter |
| Elements | ... | ... | ... | fewer |
| Max RMS spot | ... | ... | ... | smaller |
| On-axis MTF@50 | ... | ... | ... | higher |
| Edge MTF@50 | ... | ... | ... | higher |
| Glasses | ... | ... | ... | — |
### Summary
[2-3 sentences: Design A wins on X, Design B on Y. Recommendation: Design [X]
is the stronger candidate because ...]The report ends with the recommendation. It begins with the literal token
Recommendation: and names one design. Keep that marker exactly as written, so the
verdict stays easy to find for the reader and in later sessions — do not substitute prose
for it.
The Edge MTF@50 row is the one that decides most comparisons. Keep it, and keep
it honest: it is the worst field, not the mid-field.
Step 4: Restore original state
After the comparison, reload whichever design the user was working on before it started. If there was none, say so rather than leaving the end state unstated — the last design you loaded is still in the engine.
Key Rules
- Save before comparing. Loading other designs overwrites the current state.
- Same analysis on every design. Don't skip MTF on one and run it on another.
- Never compare the diffraction limit and call it on-axis. It is
get_mtfseries 0, it is nearly identical for two designs at the same f/#, and reporting it as an on-axis row turns a large real difference into a false tie. On-axis is series 1. - Never drop a field from the comparison because it is bad. Every field you analyzed appears in the report, worst included. A claim that a design is corrected "across the full field" is false unless the corner row supports it — dropping the corner and then concluding full-field correction is the failure mode this rule exists to prevent, and every number in such a report can be individually true.
- Let the user decide. Present the data, give a recommendation, but don't auto-pick — the user is the optical engineer.
- If comparing more than 4 designs, suggest narrowing down first — too many columns gets unreadable.
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
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.

