Skip to content

fix(pstack): say model tier where cross-judge still said family - #57

Merged
caelaxie merged 1 commit into
mainfrom
cursor/model-tier-wording-fc88
Nov 2, 2026
Merged

caelaxie merged 1 commit into
mainfrom
cursor/model-tier-wording-fc88

Conversation

@caelaxie

@caelaxie caelaxie commented Nov 2, 2026

Copy link
Copy Markdown
Collaborator

Closes #53

Eval, show-me-your-work, and guides 04 and 07 now say model tier / another tier, matching arena phase C.

Verification

Harness on 71a46309c4bf8ad21b68d394145157cf129f6c28 with agy 1.4.12. Proof comment follows.

Open in Web Open in Cursor 

Arena phase C already picks a judge on another tier. Match that wording in eval, show-me-your-work, and guides 04 and 07.

Co-authored-by: Kaitian Xie <caelaxie@users.noreply.github.com>
@cursor

cursor Bot commented Nov 2, 2026

Copy link
Copy Markdown

Harness proof for head 71a46309c4bf8ad21b68d394145157cf129f6c28.

  • agy 1.4.12 from the linux_amd64 manifest (sha512 prefix ea01c73b772187bc).
  • agy plugin validate --strict . passed: 61 skills, 2 agents, 0 warnings.
  • agy plugin install . from this branch installed pstack to ~/.gemini/config/plugins/pstack.
  • agy plugin list: pstack enabled, version 0.15.36, 61 skills, 2 agents (comment-sicko, poteto-agent).
  • Installed HEAD is 71a46309c4bf8ad21b68d394145157cf129f6c28, equal to this head. agy plugin validate --strict on the installed copy passed.
  • cmp against the installed copy, all identical: skills/poteto-mode/playbooks/eval.md, skills/show-me-your-work/SKILL.md, docs/guide/04-design.md, docs/guide/07-overnight.md.
  • bun test in skills/poteto-mode/scripts: 62 pass, 0 fail.
  • bun test in skills/statechart/scripts: 267 pass, 0 fail, 67 skipped (TLC tests skip without tla2tools.jar).
  • Live agy --print turns were not run. They stop at Google OAuth.

@caelaxie caelaxie left a comment

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

passes the bar. merging.

  • minimum changes: exactly the four lines #53 named (eval.md:21, show-me-your-work SKILL.md:67, guide 04:31, guide 07:68), family -> tier, nothing else touched. consistent with arena phase c and orchestrate.
  • layout/target: no cursor paths, no team-kit mention.
  • parity: version stays 0.15.36.
  • proof: agy 1.4.12 validate --strict + install + list on head 71a4630, installed copy cmp-identical for all four files; Validate plugin green. text-only, so the oauth-blocked live turn is fine per precedent.
  • linkage: Closes #53. not a draft.

@caelaxie
caelaxie merged commit bddacb8 into main Nov 2, 2026
1 check passed
@caelaxie
caelaxie deleted the cursor/model-tier-wording-fc88 branch November 2, 2026 03:11
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.

audit: cross-judge lines still say model family after the tier remap

2 participants