Repository navigation
skill: add reasoning-lens handoff and a blind-spot note - #16
Merged
Merged
Conversation
- SKILL.md: Do Not Use When gains a reasoning-lens handoff; description scoped away from single-idea examination - SKILL.md: blind-spot note for the model, not printed unless asked - CHANGELOG: record under Unreleased Stacked on fix/family-brand-and-sibling-20261006. Unevaluated: needs a regression round before merge. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01SMZASadRZRMrE9HHJn6SgC
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 changed
problem-lens/SKILL.md, "Do Not Use When": new line for requests that want one idea, claim, or plan examined from several angles. The Skill names reasoning-lens in one sentence and offers to run the options anyway.problem-lens/SKILL.md, Role: a "Blind spot" note for the model. It is not printed unless the user asks, so output length and format are unchanged.problem-lens/SKILL.md, description: now says the Skill is not for single-idea examination.CHANGELOG.md: entries under Unreleased.Why a draft: do not merge until a regression round has run
This edits
SKILL.mdafter the 1.2.0 evals, and the repo's practice is a regression round for everySKILL.mdrule (rounds 4 to 7). Suggested check:Stacked
Based on #15 (
fix/family-brand-and-sibling-20261006). Both edit the CHANGELOG's Unreleased block. Merge #15 first.Checks
python3 evals/check_rules.py problem-lens/examples/*.mdpasses. No eval round has been run on this change.🤖 Generated with Claude Code
https://claude.ai/code/session_01SMZASadRZRMrE9HHJn6SgC
Generated by Claude Code