Cross-harness parity: goose vs Claude Code vs Codex
- Dominant language
- No language data
- Stars
- 14
- Forks
- 0
- PR merge metrics
- No merged PRs in 30d
Description
The setup protocol should behave equivalently in any AGENTS.md-aware agent. Run the same Meridian dry-run (#1) in at least goose, Claude Code, and Codex, and report divergence:
- question order and phrasing drift
- phases skipped or merged
- generated file structure and frontmatter fidelity
- whether hard limits (2–4 files, ~80 lines, approval-gated rules) held
Every divergence is a protocol ambiguity to fix in `setup/PROTOCOL.md`, not a harness bug to ignore.
🤖 Seeded by @daveh-beep's agent while standing up the repo.
Contributor guide
Research direction
Start by reading setup/PROTOCOL.md and running the Meridian dry-run from issue #1 through goose, Claude Code, and Codex. Compare question order, phases, generated files and frontmatter, and the stated limits; document each divergence as a protocol ambiguity to resolve in setup/PROTOCOL.md.
Written by the indexing model from the issue text.
Assessment
- Domain
- tooling
- Issue type
- Refactor
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100