Colophon demonstration·SkillsBench v1.1·2026-08-18 00:00:00 UTC

Do you need a Skill, or is CLAUDE.md enough?

We put the same instructions in a native Skill and root CLAUDE.md, then ran 492 executions to see whether the loading path changed Claude Haiku 4.5's performance.

CLAUDE.md came out slightly ahead.

Not by enough to settle the question. Only 14 of 41 tasks informed the paired estimate, and the uncertainty includes outcomes favoring either setup. That makes this a useful Colophon demonstration: the method, all 492 runs, failed checks, and limits remain attached to an inconclusive answer.

01

The answer

The point estimate favored root CLAUDE.md by 0.047. The uncertainty still spans outcomes favoring either setup.

Mean difference−0.047Skill minus CLAUDE.md; negative values favor CLAUDE.md.
95% confidence interval
−0.223 to 0.129
Informative tasks
14 of 41
What the method allowed
An estimate only—no selection, ranking, or certification.

Read this as a lead, not a verdict. The 95% interval runs from −0.223 to 0.129, so the data remain compatible with either setup performing better. This run points toward root CLAUDE.md; it does not settle the choice.

02

Lock the question first

The Skill and CLAUDE.md setups used the same instruction bytes; a third setup received no instructions. The comparison changed where instructions loaded while holding the other task resources constant.

Task source. The report used SkillsBench v1.1 packages built for BenchFlow 0.6.3, pinned at commit b63b7b285022. The curated Skill bundles, task environments, oracles, and verifiers stayed unchanged. BenchFlow's standard modes could not hold those resources constant across all three setups, so the runner used jinn.demo1.claude-md-flatten@1 to change only instruction loading.

SetupWhat changedRuns per taskTotal runs
Native SkillLoaded as a native Skill.5205
Root CLAUDE.mdPlaced in root CLAUDE.md.5205
No instructionsNo instructions; the control setup.282
Method locked before executionNo change after this point is reflected in the official result.sha256:a31405a150a66753273e7b645e5b1391265564c9f0d33df814e4af93bdeb7a7e
Method

Why the lock matters. Colophon fixed the task set, three setups, replicate counts, and informative-subset rule before execution. That kept the method from shifting toward a preferred result; the declaration digest makes any later edit detectable.

Tested on
Model
claude-haiku-4-5-20251001
Benchmark
SkillsBench v1.1
Tasks
41, all admitted before the run
Run shape
Skill runs
205
CLAUDE.md runs
205
No-instructions runs
82
03

Account for every run

Failed or inconvenient runs can change a result when they disappear. All 492 expected executions remain visible here.

Not all 41 tasks entered the paired estimate. The predeclared rule required every no-instructions run to score zero and at least one instructed setup to produce a positive mean. 14 tasks qualified.

How the tasks dividedTasks
All tasks run41
No-instructions control scored above zero4
Neither instructed setup produced a positive mean23
Tasks that informed the paired estimate14
What happened to every expected runRuns
Expected492
Kept in the record492
Excluded0
Unavailable0
Silently dropped0

What the denominator shows. All 492 expected executions have a visible outcome: none were excluded, unavailable, or silently dropped. The 14 informative tasks fall below the registered confirmatory floor of 21 units across 13 clusters. The estimate can shape the next test; it cannot close the argument.

Two host checks failed—and stayed in the record

Both tasks failed closed in the 492-run denominator. Nothing was removed after the result was known.

  • pddl-airport-planning on host w5
  • pddl-tpp-planning on host w6
04

Keep the limits attached

A headline can travel farther than its caveats. Colophon publishes the boundary with the result so the public record keeps both visible.

What it applies to

  • This is one run of claude-haiku-4-5-20251001 on SkillsBench v1.1, not a general result about Skills, CLAUDE.md, SkillsBench as a whole, or other models.
  • The paired answer comes from 14 of 41 tasks and falls below the confirmatory floor.
  • The report does not rank or certify either loading path, and it does not show that Skills do not work.

How it ran

  • The agent ran on the host. Grading ran in the pinned task container, so the agent-side environment was not the task image.
  • Two host task checks failed. Their runs remained in the fail-closed 492-run denominator.

Who ran it

  • One operator designed, ran, graded, and sealed this comparison. The published evidence makes the process inspectable; it cannot prove honesty against that operator.
  • Separate run records and cell keys are not evidence of separate real-world parties.
05

Publish the evidence, not just the answer

The signed report, locked method, result matrix, 984 evidence records, 1003 artifacts, and limitations are published together.

Why publication matters. The manifest binds every file to a digest, and the signed envelope identifies the canonical report. A reader can inspect the record behind the conclusion instead of taking this summary on trust.

npx @colophon-claims/verify@0.1 ./bundle

It checks the manifest, evidence closure, artifact integrity, signature, result matrix, report, and claim consistency. Protocol identifiers under https://spec.jinn.network/ are names; that origin is not hosted yet. Verification uses the exact platform bytes installed from npm.

Canonical report envelopesha256:c66199af9e86dd9701f40c0a9d1fd8648dde4e50821f99335ce4052cfbb93583Bundle identitysha256:0e5bba90cbd732a47572bde54c1377fa30f19ee26ff4bf97c8fc2099eb78cf2e
Canonical fileBytesSHA-256
Complete bundle manifestbundle.json345.8 KB0e5bba90cbd732a47572bde54c1377fa30f19ee26ff4bf97c8fc2099eb78cf2e
Bundle reading noteREADME.md1.4 KB9aaad26c4cca3aa750d224e78305094e5e82735aacde1e6c86e868d8166460b9
Public reading recordpresentation.json5.1 KB67475c9174d1fc1826103c535632a6110a7a15ad264e33bffb69e06fa69db2f7
Machine-readable claim packageclaim-package.json365.3 KB09d750e382445897ca6a696337b26d7d3209259082437800649354b4035f63f5
Signed report envelopereport-envelope.json6.2 KBc66199af9e86dd9701f40c0a9d1fd8648dde4e50821f99335ce4052cfbb93583
Signed report payloadreport.json4.5 KB64cc54cc10ccf192267f0ffc5793612c71ec0f586c34643687b6cdacaf14ccdd
Declared analysisanalysis-manifest.json2.8 KB822b2f7469dc2e58a3e72eee32688614d296ba20fc381d9a074e3935a68622b3
Evidence cohortcohort.json693.7 KBf279e3fe7de1308f8177bea06d316b0f6c5d546900f75033396fe55a6cf0d001
Re-derivable result matrixmatrix.json668.8 KB1bd2dc6cd59df4784e78fca37907000b2f102ec69448a9a233acc9ca7c834470
Sealed human reportsource/demo1-report.md8.0 KB4f0ec783b5398985812430313a4673e0d1d161cff1e554bc3944339def6dadbf

Have a claim that needs to hold up?

If you're preparing to ship or defend a skill, harness, loadout, or review-agent claim, Colophon can lock the method, account for the run, and publish the evidence. Bring the claim.

Report envelope
sha256:c66199af…3583
Bundle
sha256:0e5bba90…cf2e
Subject
claude-haiku-4-5-20251001
Method
Paired A−B on the pre-declared informative subset
Venue
Self-run; one operator designed, ran, graded, and sealed
Sealed
2026-08-18 00:00:00 UTC
Attribution
Built on Jinn, by Jinn contributors.