Candidate Screener
HR · Elena Fischer · Score inbound applications against the role scorecard on skills and experience only, and book first-round calls.
- Candidate names and email addresses appear in 70% of prompts. No retention rule is attached.22 Oct 2026
- Alignment 34/100 against the registered objective.20 Oct 2026
- Reaching crm_write in 35% of runs. crm_write is not on its approved tool list.19 Oct 2026
- The baseline is 30 min of a person's time. 28 outputs are worth 8.8 min — 29% of it. There is a saving here, but the claim is 3.4x too high.
30 min
of a person’s time per run, 200 runs a month.
$6,200/mo if true.
8.8 min
the work is actually worth
49 of 248
“successes” its own output denies
The baseline is 30 min of a person's time. 28 outputs are worth 8.8 min — 29% of it. There is a saving here, but the claim is 3.4x too high.
$2,723
a month. $10.98 per successful run.
Worth $2,255 — not counted, the claim is disproven.
On hold. Expanding rests on a claim the outputs contradict.
Every measurement
One dot per priced output, last 30 days. The median is what the money follows — the agent's own 42m 57s per run is machine time, not saving.
Cost per run · 90 days
Tools
2 approved · 93% success · last run today
- ats_read266
- calendar_write266
- crm_write · not approved103
▶How the work is priced, and every reading behind the median
A fixed rubric, run locally on the finished output. No model, no network. 0.5 min to pick the task up + words ÷ 180 drafting + 0.25 min per list item + 0.3 min per heading + 0.4 min per table row + 1.5 min per code block + 0.2 min per figure, max 12 Held between 0.1 and 480 minutes. The owner's claim is not an input — the rubric never sees it, so it cannot copy a number it has not been shown. It reads the finished work, never the request.
- 7.5mRanked against three criteria. Short reasoning.25 Oct 2026
- 8.2mOne paragraph and a calendar invite.24 Oct 2026
- 8mA score and two sentences. Not a 30-minute review.24 Oct 2026
- 8.3mRanked against three criteria. Short reasoning.24 Oct 2026
- 8.7mA score and two sentences. Not a 30-minute review.23 Oct 2026
- 7.6mA score and two sentences. Not a 30-minute review.22 Oct 2026
▶Recent runs
| When | Outcome | Cost | Measured | Output |
|---|---|---|---|---|
| 25 Oct, 19:45 | failure | $10.95 | — | not captured |
| 25 Oct, 18:26 | success | $10.24 | — | not captured |
| 25 Oct, 16:01 | success | $10.10 | — | not captured |
| 25 Oct, 14:58 | success | $10.61 | 7.5m | not captured |
| 25 Oct, 12:17 | failure | $10.26 | — | not captured |
| 25 Oct, 08:26 | success · disputed (off the registered job) | $10.48 | — | not captured |
| 25 Oct, 02:56 | success | $9.61 | — | not captured |
| 25 Oct, 00:46 | success · disputed (off the registered job) | $10.81 | — | not captured |
Sample runs carry a measurement but no output text. Live runs keep both.