Skip to content

benchmarks: record Cursor skill invocation results #37

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
anusbutt merged 2 commits into from
Aug 15, 2026
Merged

benchmarks: record Cursor skill invocation results #37

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5712b00
benchmarks: record Cursor skill invocation results
anusbutt Aug 15, 2026
2412fe3
Merge remote-tracking branch 'origin/main' into benchmarks/cursor-ski…
anusbutt Aug 15, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 24 additions & 0 deletions benchmarks/results/cursor-skill-invocation-2026-08-15.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,24 @@
# Cursor skill invocation benchmark — 2026-08-15

## Conditions

- Same repository, prompt, GraphKeeper records, and Cursor environment across runs.
- Prompt: `Investigate the homepage project opacity behavior and determine whether the opacity poll is still part of the current implementation or legacy behavior.`

## Results

| Setup | Result |
| --- | --- |
| Composer 2.5 — natural invocation #1 | FAIL |
| Composer 2.5 — natural invocation #2 | FAIL |
| Composer 2.5 + explicit `/graphkeeper` | PASS |
| Grok 4.6 — natural invocation #1 | PASS |
| Grok 4.6 — natural invocation #2 | PASS |

## Finding

In this specific Cursor benchmark, automatic GraphKeeper skill invocation differed by model. Composer 2.5 did not invoke GraphKeeper automatically in 2/2 runs, while Grok 4.6 did in 2/2 runs. Explicit `/graphkeeper` activation worked with Composer 2.5.

## Product decision

No Cursor-specific adapter is justified yet. The existing skill works in Cursor; the observed difference appears to be around automatic skill selection.