skill-authoring §2: a capability you describe for a weaker executor is a claim to verify live before the doc ships - #80
Merged
Conversation
firaen22
force-pushed
the
mine-doc-capability-claim
branch
from
July 24, 2026 15:52
00648eb to
0851619
Compare
…s a claim to verify live before the doc ships
A skill/runbook that tells a weaker model a capability exists ('the engine
auto-searches every category', 'this flag does X') is executed verbatim by
the reader least able to notice it is wrong — shipping it unverified ships a
false instruction. When the behavior is new or just-patched, prove it against
a real gate, THEN finalize the doc; a scarce/expiring live session argues for
verifying FIRST, not for shipping on the theory it probably works. Positive
twin of §3's capability-negative rule. Class-distilled from a mining pass over
the owner's own sessions (no code taken).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
firaen22
force-pushed
the
mine-doc-capability-claim
branch
from
July 25, 2026 00:11
0851619 to
89f505f
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
New §2 bullet, opening from this section's existing verify-before-you-write-it rule: a skill/runbook that tells a weaker model a capability exists ("the engine auto-searches every category", "this flag does X") is executed verbatim by the reader least able to notice it's wrong. When the behavior is new or just-patched, prove it against a real gate (ground-truth-gates), THEN finalize the doc. A scarce/expiring live session argues for verifying FIRST, not shipping on the theory it probably works.
Why
Class-distilled from a mining pass over the owner's own sessions (no code taken): doc edits describing a just-patched self-heal behavior were held until a regression gate proved the behavior, with a scarce authenticated session about to expire, on exactly this reasoning.
Ships
unprobedper the covenant..github/checks.pygreen.Revision note (2026-07-24)
Force-pushed after a same-day adversarial self-audit found the original bullet largely restated this same section's opening rule ("verify every command, flag, path, and claim... before writing it down") — about 7 lines of duplication. Also dropped the "positive-direction twin of §3's capability-negative rule" framing: §3 argues stale positive claims are the lower-severity, self-exposing class (they fail loud the first time someone follows them), which is in tension with this rule's severity claim rather than a clean mirror of it. Collapsed 18→11 lines to the genuinely non-duplicated content — the time-pressure anti-rationalization and the ground-truth-gates ordering pointer — and reframed the opener to defer to the section's existing rule instead of claiming to introduce it.
🤖 Generated with Claude Code