Electricity Bench

cursor-cli + composer-2.5 (Cursor Pro)

cursor-cli harness · Cursor Pro

plan
Cursor Pro
price
allowance
52800000 reported-tokens per month
isolation
quiet-window
usage method
estimated-from-telemetry

operator Cursor Pro login (hericko.aljaz@gmail.com) shared between eb-runner and the claude host; graded runs only inside the scheduled quiet window, all recurring tasks blocked for its duration

Solves up to tier 5 on real-world@v3 - the highest difficulty rung with a solved task
12345

Tier 1: localized single-file bug · tier 3: cause in a different module than the symptom · tier 5: problems that took a human hours. A cleared rung solved every task on it; a partial rung solved some.

overall pending Two distinct suite families are required for an overall grade.

Per-suite results

C
real-world@v3
57%
plan cost per run
1.7% of Cursor Pro
latency med / p95
117231 ms / 643992 ms
reliability
100%
error rate
0%

Run on cursor-cli 2026.07.23-e383d2b, model composer-2.5, Aug 4, 2026.

One run of this suite consumes 1.7% of Cursor Pro - about 60.2 runs per month. Measured per suite run (estimated-from-telemetry); an absolute cost on this plan, not comparable with another agent's plan. Measured in a declared quiet window on a shared subscription: operator Cursor Pro login (hericko.aljaz@gmail.com) shared between eb-runner and the claude host; graded runs only inside the scheduled quiet window, all recurring tasks blocked for its duration.

Trajectory

Every scored run of this agent, oldest first. Grades compare only within a suite version, so each version gets its own line; while a suite is being retired both versions run, and the superseded one is dashed. The filled point is the run the scorecard shows today. All agents over time →

real-world · 2 runs · latest Aug 5, 2026
Capability
0%50%100%Aug 4Aug 5cursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 4, 2026 · 57% · grade Ccursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 5, 2026 · 100% · grade A0%50%100%Aug 4Aug 5cursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 4, 2026 · 57% · grade Ccursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 5, 2026 · 100% · grade A
Median latency
104581 ms111731 ms118881 msAug 4Aug 5cursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 4, 2026 · 117231 ms · grade Ccursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 5, 2026 · 106231 ms · grade A104581 ms111731 ms118881 msAug 4Aug 5cursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 4, 2026 · 117231 ms · grade Ccursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 5, 2026 · 106231 ms · grade A
Plan cost per run - a share of this agent's own plan, not a cross-agent number
0.0%0.9%1.9%Aug 4Aug 5cursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 4, 2026 · 1.7% · grade Ccursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 5, 2026 · 0.1% · grade A0.0%0.9%1.9%Aug 4Aug 5cursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 4, 2026 · 1.7% · grade Ccursor-cli + composer-2.5 (Cursor Pro) · v3 · Aug 5, 2026 · 0.1% · grade A

Aug 4, 2026: C on v3 · Aug 5, 2026: A on v3

Compare

Transcript excerpts

Outputs are shown only when they can be separated safely from the private suite. Scores and run metadata remain visible when an output is withheld.

real-world@v3
Task#TurnsLatencyOutput
b2r5h814191681 mswithheldPrivate-suite content
c8g2m41494392 mswithheldPrivate-suite content
d9b4x615117231 mswithheldPrivate-suite content
h5n2v714404865 mswithheldPrivate-suite content
h7w3j51691667 mswithheldPrivate-suite content
k4f9t714280871 mswithheldPrivate-suite content
m6t3q813112915 mswithheldPrivate-suite content
p1x9k414666909 mswithheldPrivate-suite content
q8k3j614243781 mswithheldPrivate-suite content
r3v8m515634171 mswithheldPrivate-suite content
s2j7f41492869 mswithheldPrivate-suite content
t5s8n21346650 mswithheldPrivate-suite content
w4j7q216240535 mswithheldPrivate-suite content
y3p7k11376599 mswithheldPrivate-suite content
z6q1v91589306 mswithheldPrivate-suite content