Skill or sugar pill? Twenty skills against a same-length placebo

What did Report 017 find?

What we tested. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5.
What we found. On Claude Sonnet 5.5, 11 of 16 skills beat the fake: by the case rule, each scored above the same-length placebo on at least one of three tasks and below it on none. A skill that did not beat the fake is not shown to do no better than it.
Beating the fake is not beating no skill. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings.
What it does not show. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5.

Report 017: Placebo comparison, claude-sonnet-5-5

Type Placebo comparison report

What moved a same-length placebo text

Models claude-sonnet-5-5

Headline On Claude Sonnet 5.5, 11 of 16 skills beat the fake: by the case rule, each scored above the same-length placebo on at least one of three tasks and below it on none.

Receipts 20 linked receipts

Cite this

@techreport{driftproof-017,
  title  = {Skill or sugar pill? Twenty skills against a same-length placebo},
  author = {Driftproof},
  year   = {2026},
  type   = {Placebo comparison report},
  note   = {claude-sonnet-5-5},
  url    = {https://driftproofhq.com/reports/017/}
}

Report 017. Recorded on 1 Nov 2026. Answering model Claude Sonnet 5.5; judge Claude Haiku 5.5.

Placebo comparison report. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5.

On Claude Sonnet 5.5, 11 of 16 skills beat the fake: by the case rule, each scored above the same-length placebo on at least one of three tasks and below it on none. A skill that did not beat the fake is not shown to do no better than it.

Of the 16: ABOVE_PLACEBO 11, BELOW_PLACEBO 3, NO_DIFFERENCE 0, INCONCLUSIVE 0, UNDERPOWERED 2, NOT_MEASURED 0. Beat the fake means ABOVE_PLACEBO only.

find-skills, agent-browser, hyperframes and hyperframes-registry are left out of these counts: their text tasks test the router's written response, not the delegated skill (Report 015 Amendment 1).

Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings.

About grill-me and grill-with-docs: each skill's whole text is an instruction to call another skill. These text tasks forbid tools, so each reading reflects that one-line hand-off, not the skill it calls. Both stay in the pre-registered counts.

The skill: grill-me

---
name: grill-me
description: A relentless interview to sharpen a plan or design.
disable-model-invocation: true
---

Call the Skill tool with "grilling".

Its same-length placebo

The morning light settled slowly over the hills beyond the old market town. The museum displays a loom that weavers in the district used for linen and wool.

This text ends with one space.

grill-me, the shortest skill counted in the headline (157 bytes), beside its placebo: the same length to the byte, rebuilt from the receipt's seed, length and corpus, with the SHA-256 the receipt seals. Fake my file.

Each skill

In Report 015's order. Each row is one receipt, read by the receipt reader.

find-skills vercel-labs/skills

These text tasks test the router's written response, not the delegated skill. Report 015 Amendment 1. Left out of the headline count.

Against the same-length placebo: case rule: at least one case scored below the same-length placebo

search (separated-down): with the skill 0.75 plus or minus 0.05; with the same-length placebo 0.9125 plus or minus 0.075.
missing (underpowered): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.9 plus or minus 0.11547.
scope (underpowered): with the skill 0.85 plus or minus 0.070711; with the same-length placebo 0.95 plus or minus 0.05.

Against no skill: No verdict: delegated skill not tested.

See the question and the judge's notes: with JavaScript on, each task name above opens its prompt and rubric, the kept answers and the judge's scores and reasons, one try at a time. Without it, the kept answers sit beside this row's receipt in the repository, under Receipts at the end of the report.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: No verdict: delegated skill not tested. Report 015 receipt

Receipt 1dba083778de982f27fcbfa8e3bde3fc03c91802316ba58a2137deae5ef23c30

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, find-skills (vercel-labs/skills), Report 017, 2026-11-01. On at least one task it scored below the fake. Against no skill: No verdict: delegated skill not tested. These text tasks test the router's written response, not the delegated skill. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/1dba083778de982f27fcbfa8e3bde3fc03c91802316ba58a2137deae5ef23c30/

Why a router gets no verdict: Report 015 Amendment 1

Keep this whole message together so the limits travel with the result.

grill-me mattpocock/skills

Against the same-length placebo: case rule: Not enough draws to conclude at this effect floor

plan (underpowered): with the skill 0.866667 plus or minus 0.132916; with the same-length placebo 0.791667 plus or minus 0.106849.
constraint (underpowered): with the skill 0.466667 plus or minus 0.028868; with the same-length placebo 0.58 plus or minus 0.109545.
followup (underpowered): with the skill 0.7 plus or minus 0.122474; with the same-length placebo 0.841667 plus or minus 0.08612.

Against no skill: The skill clearly hurt: 0.678 with it and 0.749 without, on 3 test tasks.

0.678 with the skill and 0.749 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

plan (underpowered): with the skill 0.866667 plus or minus 0.132916; without it 0.785714 plus or minus 0.131385.
constraint (underpowered): with the skill 0.466667 plus or minus 0.028868; without it 0.58125 plus or minus 0.116305.
followup (separated-down): with the skill 0.7 plus or minus 0.122474; without it 0.88 plus or minus 0.057009.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Too few answers to tell. Report 015 receipt

Receipt ad5aacbe99af3e9356543dc4efe0760dfcf7300aae59a5e173369eaa136e4e96

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, grill-me (mattpocock/skills), Report 017, 2026-11-01. There were too few answers, or they varied too much, to tell it from the fake. Against no skill: The skill clearly hurt: 0.678 with it and 0.749 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. About grill-me and grill-with-docs: each skill's whole text is an instruction to call another skill. These text tasks forbid tools, so each reading reflects that one-line hand-off, not the skill it calls. Both stay in the pre-registered counts. Receipt: https://driftproofhq.com/r/ad5aacbe99af3e9356543dc4efe0760dfcf7300aae59a5e173369eaa136e4e96/

Keep this whole message together so the limits travel with the result.

grill-with-docs mattpocock/skills

Against the same-length placebo: case rule: at least one case scored below the same-length placebo

decision (separated-down): with the skill 0.75 plus or minus 0; with the same-length placebo 0.9375 plus or minus 0.075.
adr (not-separated): with the skill 1 plus or minus 0; with the same-length placebo 1 plus or minus 0.
revision (underpowered): with the skill 0.966667 plus or minus 0.028868; with the same-length placebo 0.8875 plus or minus 0.085391.

Against no skill: The skill clearly hurt: 0.906 with it and 0.922 without, on 3 test tasks.

0.906 with the skill and 0.922 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

decision (separated-down): with the skill 0.75 plus or minus 0; without it 0.91 plus or minus 0.082158.
adr (not-separated): with the skill 1 plus or minus 0; without it 1 plus or minus 0.
revision (underpowered): with the skill 0.966667 plus or minus 0.028868; without it 0.857143 plus or minus 0.093223.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly hurt. Report 015 receipt

Receipt 09e0d9f8092405058e0bde94b23385ca6599266cb018f965edbe4dddedd0ab5b

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, grill-with-docs (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored below the fake. Against no skill: The skill clearly hurt: 0.906 with it and 0.922 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. About grill-me and grill-with-docs: each skill's whole text is an instruction to call another skill. These text tasks forbid tools, so each reading reflects that one-line hand-off, not the skill it calls. Both stay in the pre-registered counts. Receipt: https://driftproofhq.com/r/09e0d9f8092405058e0bde94b23385ca6599266cb018f965edbe4dddedd0ab5b/

Keep this whole message together so the limits travel with the result.

agent-browser vercel-labs/agent-browser

These text tasks test the router's written response, not the delegated skill. Report 015 Amendment 1. Left out of the headline count.

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

web (underpowered): with the skill 0.93 plus or minus 0.057009; with the same-length placebo 0.9 plus or minus 0.070711.
desktop (underpowered): with the skill 0.821429 plus or minus 0.099403; with the same-length placebo 0.7375 plus or minus 0.047871.
dashboard (separated-up): with the skill 0.816667 plus or minus 0.028868; with the same-length placebo 0.65 plus or minus 0.108012.

Against no skill: No verdict: delegated skill not tested.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: No verdict: delegated skill not tested. Report 015 receipt

Receipt 5c56836a4ec21cb3cda2ad5552f4b4bb53876b72e5a512def9b632a948488a3e

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, agent-browser (vercel-labs/agent-browser), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: No verdict: delegated skill not tested. These text tasks test the router's written response, not the delegated skill. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/5c56836a4ec21cb3cda2ad5552f4b4bb53876b72e5a512def9b632a948488a3e/

Why a router gets no verdict: Report 015 Amendment 1

Keep this whole message together so the limits travel with the result.

improve-codebase-architecture mattpocock/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

candidate (underpowered): with the skill 0.7 plus or minus 0.061237; with the same-length placebo 0.583333 plus or minus 0.112546.
adr (separated-up): with the skill 0.95 plus or minus 0; with the same-length placebo 0.7375 plus or minus 0.047871.
report (separated-up): with the skill 0.7 plus or minus 0; with the same-length placebo 0.616667 plus or minus 0.028868.

Against no skill: The skill clearly helped: 0.783 with it and 0.704 without, on 3 test tasks.

0.783 with the skill and 0.704 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

candidate (underpowered): with the skill 0.7 plus or minus 0.061237; without it 0.6125 plus or minus 0.085391.
adr (separated-up): with the skill 0.95 plus or minus 0; without it 0.75 plus or minus 0.070711.
report (underpowered): with the skill 0.7 plus or minus 0; without it 0.75 plus or minus 0.05.

The placebo against no skill, shown here and not counted: PLACEBO_BELOW_BASELINE

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Too few answers to tell. Report 015 receipt

Receipt 4fba7a9b04033dee6c13a9c615ff264cc2175f10958c0a5ba1ed2aa19d4ad401

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, improve-codebase-architecture (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.783 with it and 0.704 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/4fba7a9b04033dee6c13a9c615ff264cc2175f10958c0a5ba1ed2aa19d4ad401/

Keep this whole message together so the limits travel with the result.

tdd mattpocock/skills

Against the same-length placebo: case rule: Not enough draws to conclude at this effect floor

test (underpowered): with the skill 1 plus or minus 0; with the same-length placebo 0.983333 plus or minus 0.028868.
coupling (underpowered): with the skill 1 plus or minus 0; with the same-length placebo 0.983333 plus or minus 0.028868.
agreement (underpowered): with the skill 0.966667 plus or minus 0.028868; with the same-length placebo 0.95 plus or minus 0.05.

Against no skill: Too few answers to tell: 0.989 with the skill and 0.972 without, but the scores varied too much to call it.

Not enough draws to conclude at this effect floor. Draws needed: no draw count at these spreads (case agreement: the two arms' spreads sum to 0.058, at or above the 0.05 floor).

test (underpowered): with the skill 1 plus or minus 0; without it 0.983333 plus or minus 0.028868.
coupling (underpowered): with the skill 1 plus or minus 0; without it 0.966667 plus or minus 0.028868.
agreement (underpowered): with the skill 0.966667 plus or minus 0.028868; without it 0.966667 plus or minus 0.028868.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Too few answers to tell. Report 015 receipt

Receipt 8a54b710d73e61ff14ffb7a027676a17b57a5d8077fa08b578b37e307e16a816

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, tdd (mattpocock/skills), Report 017, 2026-11-01. There were too few answers, or they varied too much, to tell it from the fake. Against no skill: Too few answers to tell: 0.989 with the skill and 0.972 without, but the scores varied too much to call it. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/8a54b710d73e61ff14ffb7a027676a17b57a5d8077fa08b578b37e307e16a816/

Keep this whole message together so the limits travel with the result.

hyperframes-cli heygen-com/hyperframes

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

loop (separated-up): with the skill 0.95 plus or minus 0.05; with the same-length placebo 0.75 plus or minus 0.
timeline (separated-up): with the skill 0.908333 plus or minus 0.066458; with the same-length placebo 0.428571 plus or minus 0.11127.
render (separated-up): with the skill 0.9 plus or minus 0; with the same-length placebo 0.528571 plus or minus 0.13496.

Against no skill: The skill clearly helped: 0.919 with it and 0.592 without, on 3 test tasks.

0.919 with the skill and 0.592 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

loop (separated-up): with the skill 0.95 plus or minus 0.05; without it 0.7 plus or minus 0.05.
timeline (separated-up): with the skill 0.908333 plus or minus 0.066458; without it 0.592857 plus or minus 0.133631.
render (separated-up): with the skill 0.9 plus or minus 0; without it 0.483333 plus or minus 0.028868.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt bd80739ac0b870b4c7be48caf2c2423dbb691bc504314835722e6f7db64a6789

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, hyperframes-cli (heygen-com/hyperframes), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.919 with it and 0.592 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/bd80739ac0b870b4c7be48caf2c2423dbb691bc504314835722e6f7db64a6789/

Keep this whole message together so the limits travel with the result.

hyperframes heygen-com/hyperframes

These text tasks test the router's written response, not the delegated skill. Report 015 Amendment 1. Left out of the headline count.

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

existing (separated-up): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.71 plus or minus 0.054772.
usage (separated-up): with the skill 0.89 plus or minus 0.074162; with the same-length placebo 0.65 plus or minus 0.158114.
port (underpowered): with the skill 0.958333 plus or minus 0.080104; with the same-length placebo 0.93 plus or minus 0.057009.

Against no skill: No verdict: delegated skill not tested.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: No verdict: delegated skill not tested. Report 015 receipt

Receipt d517dfd71ab89a0dedf89f58b5913b0ef6f2c8eba3778f5c30100ece3d4c533f

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, hyperframes (heygen-com/hyperframes), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: No verdict: delegated skill not tested. These text tasks test the router's written response, not the delegated skill. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/d517dfd71ab89a0dedf89f58b5913b0ef6f2c8eba3778f5c30100ece3d4c533f/

Why a router gets no verdict: Report 015 Amendment 1

Keep this whole message together so the limits travel with the result.

setup-matt-pocock-skills mattpocock/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

local (separated-up): with the skill 0.93 plus or minus 0.057009; with the same-length placebo 0.65 plus or minus 0.1.
labels (underpowered): with the skill 0.8 plus or minus 0.119024; with the same-length placebo 0.642857 plus or minus 0.136713.
monorepo (underpowered): with the skill 0.95 plus or minus 0.05; with the same-length placebo 0.835714 plus or minus 0.098802.

Against no skill: The skill clearly helped: 0.893 with it and 0.671 without, on 3 test tasks.

0.893 with the skill and 0.671 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

local (separated-up): with the skill 0.93 plus or minus 0.057009; without it 0.591667 plus or minus 0.08612.
labels (underpowered): with the skill 0.8 plus or minus 0.119024; without it 0.671429 plus or minus 0.121988.
monorepo (underpowered): with the skill 0.95 plus or minus 0.05; without it 0.75 plus or minus 0.192354.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 176b26bcac810bddb1878dc76b1503232d7ccc6bff62c8e8fb022180f3487c78

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, setup-matt-pocock-skills (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.893 with it and 0.671 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/176b26bcac810bddb1878dc76b1503232d7ccc6bff62c8e8fb022180f3487c78/

Keep this whole message together so the limits travel with the result.

handoff mattpocock/skills

Against the same-length placebo: case rule: at least one case scored below the same-length placebo

resume (not-separated): with the skill 0.75 plus or minus 0; with the same-length placebo 0.75 plus or minus 0.
privacy (separated-down): with the skill 0.7375 plus or minus 0.047871; with the same-length placebo 0.95 plus or minus 0.070711.
destination (underpowered): with the skill 0.81 plus or minus 0.143178; with the same-length placebo 0.733333 plus or minus 0.093095.

Against no skill: The skill clearly hurt: 0.766 with it and 0.804 without, on 3 test tasks.

0.766 with the skill and 0.804 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

resume (underpowered): with the skill 0.75 plus or minus 0; without it 0.733333 plus or minus 0.028868.
privacy (separated-down): with the skill 0.7375 plus or minus 0.047871; without it 0.93 plus or minus 0.075829.
destination (underpowered): with the skill 0.81 plus or minus 0.143178; without it 0.75 plus or minus 0.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt c99347fd5003d4ce8f8af82defa7c3480943b1d348dc35c73845f59cf9776258

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, handoff (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored below the fake. Against no skill: The skill clearly hurt: 0.766 with it and 0.804 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/c99347fd5003d4ce8f8af82defa7c3480943b1d348dc35c73845f59cf9776258/

Keep this whole message together so the limits travel with the result.

frontend-design anthropics/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

identity (underpowered): with the skill 1 plus or minus 0; with the same-length placebo 0.983333 plus or minus 0.028868.
brief (separated-up): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.866667 plus or minus 0.028868.
content (underpowered): with the skill 0.82 plus or minus 0.09083; with the same-length placebo 0.8 plus or minus 0.108012.

Against no skill: Too few answers to tell: 0.934 with the skill and 0.936 without, but the scores varied too much to call it.

Not enough draws to conclude at this effect floor. Draws needed: no draw count at these spreads (case content: the two arms' spreads sum to 0.177, at or above the 0.05 floor).

identity (underpowered): with the skill 1 plus or minus 0; without it 0.983333 plus or minus 0.028868.
brief (underpowered): with the skill 0.983333 plus or minus 0.028868; without it 0.95 plus or minus 0.
content (underpowered): with the skill 0.82 plus or minus 0.09083; without it 0.875 plus or minus 0.086603.

The placebo against no skill, shown here and not counted: PLACEBO_BELOW_BASELINE

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Too few answers to tell. Report 015 receipt

Receipt 0868c96e9921658415dfbc23245185d69a3b867f1af4a146f0adb0d0c75cd007

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, frontend-design (anthropics/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: Too few answers to tell: 0.934 with the skill and 0.936 without, but the scores varied too much to call it. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/0868c96e9921658415dfbc23245185d69a3b867f1af4a146f0adb0d0c75cd007/

Keep this whole message together so the limits travel with the result.

reddit-automation flowkit-labs/skills

Against the same-length placebo: case rule: at least one case scored below the same-length placebo

rank (separated-down): with the skill 0.766667 plus or minus 0.028868; with the same-length placebo 0.883333 plus or minus 0.08756.
reply (underpowered): with the skill 0.766667 plus or minus 0.028868; with the same-length placebo 0.7875 plus or minus 0.075.
rules (separated-down): with the skill 0.6875 plus or minus 0.062915; with the same-length placebo 0.85 plus or minus 0.057735.

Against no skill: The skill clearly hurt: 0.740 with it and 0.810 without, on 3 test tasks.

0.740 with the skill and 0.810 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

rank (underpowered): with the skill 0.766667 plus or minus 0.028868; without it 0.858333 plus or minus 0.102062.
reply (underpowered): with the skill 0.766667 plus or minus 0.028868; without it 0.7875 plus or minus 0.047871.
rules (separated-down): with the skill 0.6875 plus or minus 0.062915; without it 0.783333 plus or minus 0.028868.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly hurt. Report 015 receipt

Receipt 634c8503a6289ce2d01baf865a4baaca5d3fe1a3d78dc16f971b86db1fa56d87

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, reddit-automation (flowkit-labs/skills), Report 017, 2026-11-01. On at least one task it scored below the fake. Against no skill: The skill clearly hurt: 0.740 with it and 0.810 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/634c8503a6289ce2d01baf865a4baaca5d3fe1a3d78dc16f971b86db1fa56d87/

Keep this whole message together so the limits travel with the result.

triage mattpocock/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

conflict (separated-up): with the skill 0.8875 plus or minus 0.075; with the same-length placebo 0.633333 plus or minus 0.028868.
evidence (underpowered): with the skill 0.933333 plus or minus 0.028868; with the same-length placebo 0.966667 plus or minus 0.028868.
ready (separated-up): with the skill 0.875 plus or minus 0.086603; with the same-length placebo 0.483333 plus or minus 0.028868.

Against no skill: The skill clearly helped: 0.899 with it and 0.631 without, on 3 test tasks.

0.899 with the skill and 0.631 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

conflict (separated-up): with the skill 0.8875 plus or minus 0.075; without it 0.5875 plus or minus 0.118145.
evidence (underpowered): with the skill 0.933333 plus or minus 0.028868; without it 0.8375 plus or minus 0.094648.
ready (separated-up): with the skill 0.875 plus or minus 0.086603; without it 0.466667 plus or minus 0.028868.

The placebo against no skill, shown here and not counted: PLACEBO_ABOVE_BASELINE

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 89b67ed69a5a7368214cac086f1ef50b57e19c4644af33d71f456663ead117e8

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, triage (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.899 with it and 0.631 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/89b67ed69a5a7368214cac086f1ef50b57e19c4644af33d71f456663ead117e8/

Keep this whole message together so the limits travel with the result.

prototype mattpocock/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

logic (separated-up): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.275 plus or minus 0.189077.
ui (separated-up): with the skill 0.916667 plus or minus 0.028868; with the same-length placebo 0.6 plus or minus 0.
unclear (separated-up): with the skill 0.7625 plus or minus 0.062915; with the same-length placebo 0.45 plus or minus 0.070711.

Against no skill: The skill clearly helped: 0.887 with it and 0.401 without, on 3 test tasks.

0.887 with the skill and 0.401 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

logic (separated-up): with the skill 0.983333 plus or minus 0.028868; without it 0.233333 plus or minus 0.163299.
ui (separated-up): with the skill 0.916667 plus or minus 0.028868; without it 0.51875 plus or minus 0.212027.
unclear (separated-up): with the skill 0.7625 plus or minus 0.062915; without it 0.45 plus or minus 0.05.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 2026a8968d1a5ae85d33c02e634987021b04edd9ba568a9e1d2436f56621ab89

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, prototype (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.887 with it and 0.401 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/2026a8968d1a5ae85d33c02e634987021b04edd9ba568a9e1d2436f56621ab89/

Keep this whole message together so the limits travel with the result.

media-use heygen-com/hyperframes

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

reuse (separated-up): with the skill 0.85 plus or minus 0.11547; with the same-length placebo 0.6 plus or minus 0.05.
logo (underpowered): with the skill 0.94 plus or minus 0.082158; with the same-length placebo 0.85 plus or minus 0.086603.
grade (underpowered): with the skill 0.9 plus or minus 0; with the same-length placebo 0.92 plus or minus 0.083666.

Against no skill: The skill clearly helped: 0.897 with it and 0.769 without, on 3 test tasks.

0.897 with the skill and 0.769 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

reuse (separated-up): with the skill 0.85 plus or minus 0.11547; without it 0.607143 plus or minus 0.113389.
logo (underpowered): with the skill 0.94 plus or minus 0.082158; without it 0.81 plus or minus 0.119373.
grade (underpowered): with the skill 0.9 plus or minus 0; without it 0.89 plus or minus 0.074162.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 9b55433e219fb065e304806bceea6aee214d3b3fb55eaa11244eef5ef5533586

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, media-use (heygen-com/hyperframes), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.897 with it and 0.769 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/9b55433e219fb065e304806bceea6aee214d3b3fb55eaa11244eef5ef5533586/

Keep this whole message together so the limits travel with the result.

grilling mattpocock/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

frontier (separated-up): with the skill 0.8375 plus or minus 0.062915; with the same-length placebo 0.733333 plus or minus 0.028868.
settled (separated-up): with the skill 0.7 plus or minus 0.05; with the same-length placebo 0.583333 plus or minus 0.028868.
finish (underpowered): with the skill 0.433333 plus or minus 0.08165; with the same-length placebo 0.3125 plus or minus 0.047871.

Against no skill: The skill clearly helped on 1 of 3 test tasks: the frontier task decided it. Over all 3, the averages were 0.657 with it and 0.670 without.

0.657 with the skill and 0.670 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

frontier (separated-up): with the skill 0.8375 plus or minus 0.062915; without it 0.75 plus or minus 0.
settled (underpowered): with the skill 0.7 plus or minus 0.05; without it 0.7 plus or minus 0.089443.
finish (underpowered): with the skill 0.433333 plus or minus 0.08165; without it 0.56 plus or minus 0.124499.

The placebo against no skill, shown here and not counted: PLACEBO_BELOW_BASELINE

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 667515dbf9bc6f239ba757bf64e0f301488eb7dfe30d2ce1bbb6c3231890a6e3

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, grilling (mattpocock/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped on 1 of 3 test tasks: the frontier task decided it. Over all 3, the averages were 0.657 with it and 0.670 without. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/667515dbf9bc6f239ba757bf64e0f301488eb7dfe30d2ce1bbb6c3231890a6e3/

Keep this whole message together so the limits travel with the result.

hyperframes-animation heygen-com/hyperframes

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

timeline (separated-up): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.828571 plus or minus 0.118523.
runtime (separated-up): with the skill 0.7 plus or minus 0.093541; with the same-length placebo 0.1375 plus or minus 0.085391.
template (separated-up): with the skill 0.933333 plus or minus 0.028868; with the same-length placebo 0.6 plus or minus 0.05.

Against no skill: The skill clearly helped: 0.872 with it and 0.556 without, on 3 test tasks.

0.872 with the skill and 0.556 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

timeline (underpowered): with the skill 0.983333 plus or minus 0.028868; without it 0.84 plus or minus 0.124499.
runtime (separated-up): with the skill 0.7 plus or minus 0.093541; without it 0.266667 plus or minus 0.132916.
template (separated-up): with the skill 0.933333 plus or minus 0.028868; without it 0.5625 plus or minus 0.075.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 932a427a39bc105f634287f8af286d9ff5c6f58665aa3d719321ca00efaa8a4b

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, hyperframes-animation (heygen-com/hyperframes), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.872 with it and 0.556 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/932a427a39bc105f634287f8af286d9ff5c6f58665aa3d719321ca00efaa8a4b/

Keep this whole message together so the limits travel with the result.

hyperframes-registry heygen-com/hyperframes

These text tasks test the router's written response, not the delegated skill. Report 015 Amendment 1. Left out of the headline count.

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

block (separated-up): with the skill 0.9 plus or minus 0.070711; with the same-length placebo 0.5 plus or minus 0.
component (separated-up): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.294444 plus or minus 0.200693.
search (separated-up): with the skill 1 plus or minus 0; with the same-length placebo 0.4875 plus or minus 0.085391.

Against no skill: No verdict: delegated skill not tested.

The placebo against no skill, shown here and not counted: PLACEBO_ABOVE_BASELINE

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: No verdict: delegated skill not tested. Report 015 receipt

Receipt 5fc27143c0cb1bc36e6915a4f162edec66f29a89168e7c75c02ccd251248d2bf

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, hyperframes-registry (heygen-com/hyperframes), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: No verdict: delegated skill not tested. These text tasks test the router's written response, not the delegated skill. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/5fc27143c0cb1bc36e6915a4f162edec66f29a89168e7c75c02ccd251248d2bf/

Why a router gets no verdict: Report 015 Amendment 1

Keep this whole message together so the limits travel with the result.

design-mobile-apps designed-by-ai/skills

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

response (underpowered): with the skill 0.85 plus or minus 0.173205; with the same-length placebo 0.683333 plus or minus 0.08165.
shots (separated-up): with the skill 0.92 plus or minus 0.075829; with the same-length placebo 0.683333 plus or minus 0.028868.
ids (underpowered): with the skill 0.7 plus or minus 0; with the same-length placebo 0.6125 plus or minus 0.103078.

Against no skill: The skill clearly helped: 0.823 with it and 0.739 without, on 3 test tasks.

0.823 with the skill and 0.739 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

response (underpowered): with the skill 0.85 plus or minus 0.173205; without it 0.766667 plus or minus 0.121106.
shots (separated-up): with the skill 0.92 plus or minus 0.075829; without it 0.733333 plus or minus 0.028868.
ids (underpowered): with the skill 0.7 plus or minus 0; without it 0.716667 plus or minus 0.028868.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 9fa7bc62d8703ea066da8b40f0b3adc7c51a153f524155767b5c9e24f4042c1e

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, design-mobile-apps (designed-by-ai/skills), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.823 with it and 0.739 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/9fa7bc62d8703ea066da8b40f0b3adc7c51a153f524155767b5c9e24f4042c1e/

Keep this whole message together so the limits travel with the result.

hyperframes-core heygen-com/hyperframes

Against the same-length placebo: case rule: at least one case scored above the same-length placebo, and none below

root (underpowered): with the skill 0.983333 plus or minus 0.028868; with the same-length placebo 0.93 plus or minus 0.067082.
transform (separated-up): with the skill 0.95 plus or minus 0.05; with the same-length placebo 0.75 plus or minus 0.
exit (separated-up): with the skill 0.81 plus or minus 0.089443; with the same-length placebo 0.366667 plus or minus 0.028868.

Against no skill: The skill clearly helped: 0.914 with it and 0.651 without, on 3 test tasks.

0.914 with the skill and 0.651 without; on at least one of the 3 test tasks, the two ranges of scores (the average plus or minus one spread) did not overlap, and the gap was at least 0.05.

root (underpowered): with the skill 0.983333 plus or minus 0.028868; without it 0.9125 plus or minus 0.118145.
transform (separated-up): with the skill 0.95 plus or minus 0.05; without it 0.7 plus or minus 0.05.
exit (separated-up): with the skill 0.81 plus or minus 0.089443; without it 0.34 plus or minus 0.155724.

The placebo against no skill, shown here and not counted: UNDERPOWERED

Context only, not this measurement: Report 015 read this skill in a separate, earlier run on 2026-10-18: Clearly helped. Report 015 receipt

Receipt 7c7564b040f77c1bb69904c78545d55c1d8dc08b8661b3ef5c9522c0de9a6245

Preview and copy this result

The finding, with its limits

View card image

On Claude Sonnet 5.5, judged by Claude Haiku 5.5, hyperframes-core (heygen-com/hyperframes), Report 017, 2026-11-01. On at least one task it scored above the fake, and on none below. Against no skill: The skill clearly helped: 0.914 with it and 0.651 without, on 3 test tasks. Report 015's twenty skills on Sonnet 5.5, each against no skill and against a same-length text from a fixed corpus, judged by Haiku 5.5. Where the fake scored below no skill, beating it does not mean beating no skill; each row shows both readings. It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5. Receipt: https://driftproofhq.com/r/7c7564b040f77c1bb69904c78545d55c1d8dc08b8661b3ef5c9522c0de9a6245/

Keep this whole message together so the limits travel with the result.

Method

These sections are the pre-registered note, as the campaign first stored it.

What each receipt shows

Each of the 20 frozen attempts runs one skill of Report 015's plain list on claude-sonnet-5-5, graded by claude-haiku-5-5, with three arms: the skill text, no skill, and a text of the skill's length in bytes from the revision 1 fixed corpus (spec 440).

The report shows two readings for each receipt, each in lib/verdict.js's own words:

the verdict: receiptVerdict, the skill against no skill, in the receipt reader's words;
the placebo reading: placeboLine(placeboReading(receipt)), the skill against the same-length text.

The verdict shown is the Report 017 receipt's own. Report 015's verdicts are not put beside it as if they were the same run. The verdict reads only the skill and no-skill arms, so the placebo arm's replies move it in no way. No figure other than these readings and their counts enters the headline.

Which receipt is read

The report reads exactly one sealed receipt per attempt id. It is the receipt the campaign's manifest chain records for that attempt, read through the campaign ledger. The first completed receipt is the only one: an attempt with a sealed receipt is never run again, and a second receipt for the same attempt id refuses the whole chain. No receipt is chosen after its results are seen.

An attempt without a receipt may be run again in a later invocation of the same frozen plan, at most three runs in all, its earlier costs counted against the USD 400 cap. Every failure stays in the ledger. A failure carries no reading, so running it again chooses nothing.

The campaign has one root directory. It keeps the plan and this note as first started and an append-only log with one line per start. A second start with no prior chain is refused, and a resume must continue the log's latest manifest. The report reads only the chain that ends at the log's latest manifest, and refuses if the root holds any other manifest under this plan. A cost correction may be added after a run and is never rewritten: each manifest records the corrections in force when it started.

Terms

"Beat the fake" means exactly one thing: the placebo reading is ABOVE_PLACEBO. In the reader's words, "case rule: at least one case scored above the same-length placebo, and none below". The case rule reads each of the three tasks by its draw bands, so this says no more than that, on at least one of three tasks, the skill's band sat above the placebo's at the effect floor, and on none below.

No other reading beats the fake. BELOW_PLACEBO, NO_DIFFERENCE, INCONCLUSIVE, UNDERPOWERED and NOT_MEASURED do not. None of them shows that a skill does no better than the fake.

A missing receipt reads NOT_MEASURED: a failed, cancelled or budget-stopped attempt is counted as not read, never as a result. A stub receipt also reads NOT_MEASURED.

Who is counted

The four routers are find-skills, agent-browser, hyperframes and hyperframes-registry. They carry Report 015 Amendment 1's qualification: their text tasks test the router's written response, not the delegated skill. Their readings are shown with that qualification. They are left out of the headline count, as Report 015 leaves them out of its verdict counts.

The headline counts the other 16 skills. N is 16.

The secondary reading

Beside each skill, and never in the headline, the report shows a second reading: the same case rule over the placebo's rows against the no-skill rows of the same receipt (tally.cjs secondaryReading). Its tokens are PLACEBO_ABOVE_BASELINE, PLACEBO_BELOW_BASELINE, NO_DIFFERENCE, INCONCLUSIVE, UNDERPOWERED and NOT_MEASURED. It is read only where the placebo reading is read, and refused wherever that is refused. It is not counted, not subtracted from the placebo reading, and not combined with it.

Limits

It does not show how these skills do beyond three text tasks each, written by Driftproof for Report 015's twenty skills from the Skills.sh snapshot of 2026-10-17, on Claude Sonnet 5.5 judged by Claude Haiku 5.5.

Three text-only tasks. Driftproof wrote the suite from the pinned skill description and README. These are not author-supplied tests. External tools, generated media and deployed services were not tested.

The eval account's configuration was read and found clean before and after each run, not during it.

For find-skills, agent-browser, hyperframes and hyperframes-registry: This tests the router's written response; it does not test the delegated skill.

The placebo text is not inert: on a task that supplies the facts its answer needs, a same-length text about landscape and village life can pull an answer off them, so the fake can score below no skill. The secondary reading shows where it did.

It does not say why a skill helps, that a skill and the fake are the same, or that the fake carries no information about a task. It says nothing about a skill that was not run with the placebo arm. It does not combine the 20 readings into one score.

The placebo corpus was chosen without reference to any task, so a placebo text can still bear on a task by chance.

Unkeyed seals and retained outputs permit checking internal consistency; they do not authenticate an origin.

Run record

One campaign root, receipts/report-017c, read from the latest manifest its log names. 20 of 20 attempts have a sealed receipt; every other attempt is a recorded failure and reads NOT_MEASURED. The earliest receipt date is 2026-11-01 and the latest 2026-11-01. No finish time is inferred.

The surface was claude-cli. Recorded spend across the chain: USD 22.712233. Recorded costs are API-equivalent estimates, not a subscription invoice.

receipts/report-017c/campaign-1/manifest.json, SHA-256 b87f0c7ca4f22d4c2278e71c35c400d06f01356f16b3faca5e627ac4b953e318

Plan SHA-256 3f008b21d542b85b90a5e37c60672c62788b9742319ca1f1fde8c906c419f3f4. Pre-registered note SHA-256 404b6629b6a7e0e376e10f15926392ddfa6fdf011c4df5dfd5eac9786dcd6c84.

Read and built by:

scripts/prepare-report-017.js, SHA-256 b61c1f00f86bb5b38d869324da697ea8dad5cd61b41a8b6597a0391b75a992a3
specs/613-placebo-skills-report/publication.cjs, SHA-256 fba3577194bab75b0ac4c993b4567008c418e17178ef4399e560a71ff2754050
specs/613-placebo-skills-report/tally.cjs, SHA-256 a7d334796f32591a9488baa2286e3ba9462305ff6a92b8b2f0c7695a96a415dd
specs/613-placebo-skills-report/run-campaign.cjs, SHA-256 db070608736eb9401fb8acf6a15cf800deeefbd706b18a92383242ef6609af6a
specs/408-skill-measurements/ledger.cjs, SHA-256 263a3d3423feea70aaf90074a58bcaf48d4ddb8b0eea262ec634450b17b63aae
specs/408-skill-measurements/reconstruct.cjs, SHA-256 f64871c647c3482a0d05e6a2308a05192b5934816097978ece34d553a360f2dd
lib/verdict.js, SHA-256 336cacca9ab4f80797de2ad12da8b393984843863f37e6839929aa8d44859ce8
lib/plain.js, SHA-256 46f471b1487460b6101680f0b4c1191e358611020f326898e37f0e420011dbb9
lib/placebo.js, SHA-256 f9511bdbb9928f615017addebfa305b1f7b27c544cc5dbd377e668d661be26a0
lib/receipt.js, SHA-256 6f4d56e9be11dd8dea52810d458d4c84b3eb188359f834e3bc771f2f3f8605d8
scripts/skill-handoff.js, SHA-256 261c15ba3c979f74bf69d4a6f7e192f8458a0b9635b7a17effd8e7cf1d53fda2

Amendments

Amendment 1, 2 Nov 2026

Added a note after the headline and in the rows for grill-me and grill-with-docs. The whole text of each skill only calls another skill, and these text tasks forbid tools. Counts, readings and receipts are unchanged.

Amendment 2, 2 Nov 2026

Added a link, after the opening panel, to a blind read of handoff's kept answers. Counts, readings and receipts are unchanged.

Amendment 3, 3 Nov 2026

Reworded the line against no skill in the rows where the overall averages with and without the skill do not point the verdict's way: it names the task that decided it and gives the averages over all the tasks. The verdict is decided per task, so those averages can sit equal or point the other way. Counts, readings and receipts are unchanged.

Amendment 4, 4 Nov 2026

Added, under each task of every row with a receipt, a way to open that task's prompt and rubric, the kept answers with the skill, with the fake text and with nothing, and the judge's scores and reasons for one try at a time. Counts, readings and receipts are unchanged.

Amendment 5, 4 Nov 2026

Added Send this result to each row, with a fixed message that keeps its reading, model, judge, report, caveats and receipt together. A router row also links Report 015 Amendment 1, which says why a router gets no verdict. Counts, readings and receipts are unchanged.