Skip to content

fix(writing-skills): replace count anchors with standard-based language#1692

Open
citizen204 wants to merge 1 commit into
obra:devfrom
citizen204:fix/writing-skills-count-anchor-1487
Open

fix(writing-skills): replace count anchors with standard-based language#1692
citizen204 wants to merge 1 commit into
obra:devfrom
citizen204:fix/writing-skills-count-anchor-1487

Conversation

@citizen204

Copy link
Copy Markdown

Summary

Fixes #1487, filed by @obra.

Replaces the '3+ combined pressures' count targets with qualitative standards at all 5 sites across \SKILL.md\ and \ esting-skills-with-subagents.md.

The problem

The count anchor was doing gating work that the actual standard should carry. As demonstrated in the issue, agents hit the count and stop — leaving material unexamined — rather than continuing until the standard is met. A scenario with two strong, well-aligned pressures can trigger the target rationalization more reliably than four weak unaligned ones.

Changes

*\skills/writing-skills/SKILL.md* (1 site)

  • Line 601: Replaced (3+ combined pressures for discipline skills)\ with standard-based language: scenarios that reliably trigger the rationalizations you want to prevent

*\skills/writing-skills/testing-skills-with-subagents.md* (4 sites)

  • Line 51: Removed count anchor from the RED phase checklist item
  • Line 140: Replaced \Best tests combine 3+ pressures\ with the qualitative diagnostic: does an agent without the skill consistently take the rationalization path?
  • Line 313: Removed count anchor from the final testing checklist
  • Line 344: Replaced the fix prescription \Combine 3+ pressures\ with the actual standard

The replacement standard (from the issue)

Pressure scenarios should reliably trigger the rationalizations you want to prevent. The diagnostic for whether a scenario is good enough: does an agent without the skill, run on this scenario, consistently take the rationalization path you're trying to close?

Replace '3+ combined pressures' count targets with qualitative standards
throughout SKILL.md and testing-skills-with-subagents.md (5 sites).

The count was doing gating work the actual standard could carry alone:
'scenarios that reliably trigger the rationalizations you want to prevent.'
A scenario with two strong, well-aligned pressures can trigger the
rationalization more reliably than four weak unaligned ones.

Closes obra#1487
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant