Does Skill builder help? Tested on writing skills tasks

Skill builder, from yusufkaraaslan/Skill_Seekers. Its skill authoring tasks were run with the skill loaded and without it. Each time it is the same task, run twice.

What we tested

Whether loading this skill helps a model write a skill document.

What counts as helping

The score with the skill has to beat the score without it by at least 10 percentage points, on the same 10 briefs.

How we scored it

We checked each document against the open Agent Skills specification and counted the rules it passed. The score is the share it passed, from 0 to 1.

Without to with, per model

Via API tested via API

Gemini 3.1 Flash Lite

0.745
1.000

Holds

GPT-5 mini

0.709
0.982

Holds

deterministic pass rate, 0 to 1

In Claude Code tested in Claude Code

Claude Fable 5.1 Wider run, three ways

1.000 without and with

Could not measure

Claude Haiku 4.5 Wider run, three ways

0.600
1.000

Holds

Claude Opus 5 Wider run, three ways

1.000 without and with

Could not measure

Claude Sonnet 5 Wider run, three ways

1.000 without and with

Could not measure

deterministic pass rate, 0 to 1

A picture of the per-model numbers, drawn from the same results. The tables are the source. Grey is the score without. Colour is the score with. The bracket shows how much the difference could move if we ran it again. Rows are ordered by the score with. A model measured at two versions keeps its versions next to each other. The two test methods are reported separately and never averaged. Where the two scores are the same, the number is printed once at the end of the pair. A bracket wider than the axis is drawn to the edge with its cap omitted; the table gives its bounds.
Show per-task detail

Every item, without and with, per model

Via API tested via API

  • GPT-5 mini
  • Gemini 3.1 Flash Lite

deterministic pass rate, 0 to 1

Hover or focus an item to read its task and every model’s two values.

Task-paired: these item means are what this cell's interval is built from.

The item means behind this plot
Per-item arm means. Without is the unaided arm, with is the treated arm.
ItemGPT-5 mini Gemini 3.1 Flash Lite
Task without with without with
sa-04 0.273 1.000 0.909 1.000
sa-06 0.727 1.000 0.545 1.000
sa-08 0.909 1.000 0.364 1.000
sa-10 0.727 1.000 0.545 1.000
sa-05 0.818 1.000 0.545 1.000
sa-07 0.545 1.000 0.909 1.000
sa-01 0.727 1.000 0.909 1.000
sa-02 0.727 0.909 0.909 1.000
sa-09 0.727 1.000 0.909 1.000
sa-03 0.909 0.909 0.909 1.000
A picture of the per-item means behind the per-model numbers. The tables are the source. Each model draws two lines over the same task set: a dashed line through its unaided scores and a solid line through its treated ones. Items are ordered by the mean unaided score across the models that measured them, lowest first. The two instruments are reported separately and never averaged.

Without to with, per model

In Claude Code tested in Claude Code

Claude Haiku 4.5 Expansion cohort (three arms)

Holds

Claude Fable 5.1 Expansion cohort (three arms)

Could not measure

Claude Opus 5 Expansion cohort (three arms)

Could not measure

Claude Sonnet 5 Expansion cohort (three arms)

Could not measure

deterministic pass rate, 0 to 1

A picture of the per-model numbers, drawn from the same results. The tables are the source. Each row runs from the score without to the score with. The thin bar beneath it is the bracket, and it shows how much the difference could move if we ran it again. The two test methods are reported separately and never averaged. A bracket wider than the axis is drawn to the edge with its end cap omitted; the table gives its bounds.

The Claude Code runs committed their cells as arm means and intervals rather than as per-item values, so this panel is drawn without to with rather than item by item.

What exactly was tested, and how it was scored
Repository
yusufkaraaslan/Skill_Seekers
Path
distribution/claude-plugin/skills/skill-builder
Commit
f3972efa33fa79634b96936acf1fac321cdcf7c1
Content hash
42c7e6285ae80425be3de0f00e0f7caa7dbe93ff043af34595d76be7e3a5da8f
Date tested
not yet tested

The pin is the whole of this skill's identity here. It resolves at https://github.com/yusufkaraaslan/Skill_Seekers/tree/f3972efa33fa79634b96936acf1fac321cdcf7c1/distribution/claude-plugin/skills/skill-builder, and the content hash is a sha256 over exactly the text the model was given with the skill loaded. Nothing else about the skill appears on this site.

S13-yusufkaraaslan-skill-builder

Loading distribution/claude-plugin/skills/skill-builder from yusufkaraaslan/Skill_Seekers improves outputs on skill authoring tasks.

Pass criterion
With-arm mean conformance to the open Agent Skills specification exceeds the without-arm by at least 0.10 on the same 10 briefs at the same model and settings. Key B, the house standard, is reported beside this figure and takes no part in it.
Scale
unit. Same scorer, same instrument and same scale as S02-skill-creator. Added 2026-08-31 by the source expansion under R23; adding a row to an instrument does not touch the instrument.
Task pairs planned per model
10
Notes
Shares the skill-authoring task set with every other claim in this class, so all of its rows are measured on identical items against one answer key. The slot this claim scores was pinned by the 2026-08-31 source expansion under R23 and NO MODEL HAS BEEN RUN AGAINST IT. The claim exists so the class holds one claim per pinned slot, which is what makes the row comparable the day it is run.

Injected context tokens

Injected context tokens, per arm
ArmContext charactersInjected context tokens
without00
with4,4281,039

The character count is exact: it is the length of the text the with arm is given, and the content hash above is a sha256 over that same text. The token figure is an estimate at 4.26 characters per token, the ratio the phase 1 run measured over 3,120 calls, and it is labelled an estimate until a run reports its own token counts. The without arm is given the identical prompt and nothing else, so its zero is a measurement rather than a missing value.

Verdict per model

Effect per model
ModelReadingOrbitWithout -> withEffect95 percent intervalTask pairsModel version returned
Gemini 3.1 Flash LiteDeterministic, scored against a committed answer keyStable0.7455 -> 1.0000+0.2545[0.1196, 0.3895]10 pairs, 30 of 30 recordsgemini-3.1-flash-lite
GPT-5 miniDeterministic, scored against a committed answer keyStable (10 of 10 pairs)0.7091 -> 0.9818+0.2727[0.1481, 0.3973]10 pairs, 30 of 37 recordsgpt-5-mini-2025-08-07
  • Stable: the treatment arm outscored the control arm by more than the threshold, and the interval excludes zero.

Cost

Defined in the metrics canon. Cost per task divides every dollar spent on an arm by the tasks attempted on it, including tasks whose call returned nothing, because a call that returned nothing was still billed.

Where the with arm did not win

GPT-5 mini, Deterministic, scored against a committed answer key

Drew on 1 of 10 pairs: sa-03.

In Claude Code

These cells are tested in Claude Code, on a subscription path with no API key. They are a second instrument: no figure here is averaged with one tested via API above. How the two were compared.

  • Claude Fable 5.1 · could not be measured in the three-arm cohort
  • Claude Haiku 4.5 · helped in the three-arm cohort
  • Claude Opus 5 · could not be measured in the three-arm cohort
  • Claude Sonnet 5 · could not be measured in the three-arm cohort

Expansion cohort (three arms)

one pass over each committed item. Matrix hash 0078cfa8. Pre-registration · Run report.

One row per model. Every figure is read from this run’s committed cells.
ModelNo skillWith skillDeltaIntervalPairsOrbitInjected tokens
Claude Fable 5.11.00001.0000+0.0000.000 to 0.00010Unobservablenot measured: this run records the dose per arm, in the doses list
Claude Haiku 4.50.60001.0000+0.4000.293 to 0.50710Stablenot measured: this run records the dose per arm, in the doses list
Claude Opus 51.00001.0000+0.0000.000 to 0.00010Unobservablenot measured: this run records the dose per arm, in the doses list
Claude Sonnet 51.00001.0000+0.0000.000 to 0.00010Unobservablenot measured: this run records the dose per arm, in the doses list

Key A: the open Agent Skills specification

This class is ranked on conformance to the open specification and on nothing else. The specification is published by neither cohort, it is given to neither arm, and it is pinned like any other artifact: agentskills/agentskills docs at 69ef37e9424c0a7ea9dd2293b559e43ec8176379, content hash b9079c0c10b7930e8c6a20ff2bc10cda2a3343c55185120e3f1116a1a529b220.

Key A predicates, each with the passage it comes from
PredicateWhat it requires, in our wordsWhere the specification says itScored when
frontmatter-presentThe document opens with a YAML frontmatter block, and Markdown follows it.SKILL.md format, opening sentencealways
frontmatter-name-presentA name field is present.SKILL.md format, frontmatter table, name rowalways
frontmatter-description-presentA description field is present.SKILL.md format, frontmatter table, description rowalways
name-within-length-limitThe name is between one and sixty four characters.name field, first bulletalways
name-characters-legalThe name uses only lowercase letters, digits and hyphens.name field, second bulletalways
name-no-edge-hyphenThe name neither opens nor closes on a hyphen.name field, third bulletalways
name-no-consecutive-hyphensThe name carries no run of two hyphens.name field, fourth bulletalways
description-within-length-limitThe description is between one and one thousand and twenty four characters.description field, first bulletalways
body-presentInstructions follow the frontmatter rather than an empty document.Body content, opening sentencealways
within-line-limitThe document stays under the stated line ceiling for a main SKILL.md.Progressive disclosure, closing linealways
file-references-one-level-deepEvery relative path the document points at sits at most one directory down.File references, closing linealways
compatibility-within-length-limitIf compatibility is given, it stays inside its stated ceiling.compatibility field, first bulletthe field is offered
metadata-is-string-mapIf metadata is given, it maps string keys to string values.metadata field, first bulletthe field is offered
allowed-tools-is-one-stringIf allowed-tools is given, it is a single space separated string.allowed-tools field, first bulletthe field is offered

A predicate for an optional field a document does not offer is not applicable and leaves the denominator, so a document is never credited for a rule it did not engage.

Key B: the operator's house standard, not ranked

No records exist, so Key B has no figures either. It is described here rather than left out, because a reader deciding what the ranked number means needs to know what the unranked one is.

No interval, no orbit and no verdict appears above, and none can: these are two arm means and the frozen rule is never called on them.

Every figure above is computed at build time from harness/results/runs-skills.jsonl and harness/results/ledger-skills.jsonl, both committed, by the frozen status_v1 rule and the metrics_v1 cost definitions. How a claim gets tested.

720 run records behind this page. Every verdict is computed at build time by the same frozen status_v1 rule that decides every other verdict on this site, and nothing here is written by hand.