Co-author trailer still says "Claude Opus 4.7" — bump to 4.8?
- Dominant language
- TypeScript
- Stars
- 133k
- Forks
- 19.9k
- Avg merge
- 18h 46m
- Merged PRs (30d)
- 26
Description
### Summary
The co-author trailer emitted into generated `SKILL.md` files still attributes `Claude Opus 4.7`. Now that Opus 4.8 is the current model, you may want to bump it — filing as a suggestion since it touches model-pinned infrastructure that's clearly maintained deliberately.
### Where it comes from
- **Canonical value:** `hosts/claude.ts:41`
```ts
coAuthorTrailer: 'Co-Authored-By: Claude Opus 4.7 ',
```
- **Fallback default:** `scripts/resolvers/utility.ts:372` (only used when a host doesn't set `coAuthorTrailer`)
### Generated output affected
The trailer currently renders as `Claude Opus 4.7` in:
- `document-generate/SKILL.md`
- `document-release/SKILL.md`
- `ship/SKILL.md`
### Why I'm filing instead of PRing
A naive bump isn't a clean one-liner — it intersects deliberately model-pinned pieces:
- `test/fixtures/golden/claude-ship-SKILL.md` hard-codes the 4.7 trailer, so the golden test would fail until regenerated.
- There's a whole `model-overlays/opus-4-7.md` + ~21 eval test files pinned to `claude-opus-4-7`. Those are a separate, larger migration concern and **out of scope** for this trailer bump — calling them out only so the trailer change isn't mistaken for a full model migration.
### Suggested minimal change
1. `hosts/claude.ts:41` → `Claude Opus 4.8`
2. `scripts/resolvers/utility.ts:372` fallback → `Claude Opus 4.8` (for consistency)
3. `bun run gen:skill-docs` (+ `--host codex`) to regenerate
4. Refresh `test/fixtures/golden/claude-ship-SKILL.md` to match
Happy to open a PR for just this scoped change if it'd be useful — wanted to check it's wanted first given the model-pinning around it.
Contributor guide
Research direction
Start with hosts/claude.ts:41 and the fallback in scripts/resolvers/utility.ts:372, then run bun run gen:skill-docs --host codex. Update the generated SKILL.md outputs and test/fixtures/golden/claude-ship-SKILL.md, and confirm the golden test passes while leaving the separate model overlays and eval files unchanged.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- typescript
- Domain
- documentation, tooling
- Issue type
- Refactor
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Quiet
- Clarity
- Clearly specified
- Newbie friendliness
- 54/100