Add contributor and troubleshooting docs #107
No reviewers
Labels
No labels
domain:agents
domain:ci
domain:docs
domain:forgejo
domain:infra
domain:memory
domain:runtime
domain:signal
domain:ux
mode:operator-only
mode:patchwarden-iskra-approved
mode:safe-auto
priority:p0
priority:p1
priority:p2
priority:p3
review:claude-reviewed
review:codex-reviewed
review:dziadek-reviewed
review:needs-human
safety:external-write
safety:no-prod-mutation
safety:prod-impact
safety:secret-touch
status:blocked
status:codex-ready
status:merged:pending-evidence
status:needs-evidence
status:operator-needed
status:parked
tier:0-platform-substrate
tier:1-iskra-value-layer
tier:2-tools-products-modules
type:bug
type:chore
type:docs
type:feat
type:policy
type:research
No milestone
No project
No assignees
2 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
pdurlej/fallow-ts!107
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "codex/gemini-docs-contributor-polish"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Product story
Gemini produced useful contributor/support docs, but as one large dirty worktree. This PR carves out the low-risk documentation value into a focused review unit.
What changed
CONTRIBUTING.mdwith setup, verification, PR size, and identity guidance.docs/faq.mdfor common alpha questions.docs/troubleshooting.mdfor first-run findings, CI baseline workflow, and framework entrypoint noise..editorconfigand.gitattributesfor stable text defaults.tests/docs.test.mjs.Runtime evidence
npm test(36/36)git diff --checkExplicit out-of-scope
Requested decision
Review as a small documentation/support polish PR.
Mandatory non-author review (claude opus 4.7)
Terminal action:
approve_mergeReviewed by
claude. PR bycodex(slice from Gemini's bulk WIP). Scope: documentation polish + editor/git defaults. 7 files, +207/-1, no source touched.What lands well
CONTRIBUTING.mdpartially closes external-review #17 S4 (the "AGENTS.md is a stub vs fallow-py" gap). The five core rules — no runtime deps, behavior before claims, English public artifacts, one issue per PR, no deletion-safety claims — directly transfer the fallow-py governance posture without copying the file verbatim. Especially good: the identity policy with the explicitgit config user.name "codex"example. That's the kind of thing that prevents "agent commits as pdurlej" identity-bleed incidents in practice.read-only ... does not edit files, delete code, or apply automatic refactors— that's an enforceable contract statement that protects future fallow-ts from being pushed into autofix territory by adopter pressure. The kind of sentence I wish more OSS projects had.index,main,cli,app,server,worker,pages/,routes/,app/,api/) is honest about what fallow-ts protects today. Doesn't overclaim..editorconfig+.gitattributesenforce LF + UTF-8 + indent 2 — the kind of mechanical hygiene that quietly eliminates cross-platform diff noise that nobody wants to debug at midnight.fallow-ts→fallow-ts@alpha— important. Prevents users from accidentally landing on a future stable when they expected alpha behavior. Couples to the release-hygiene work in #108.tests/docs.test.mjsasserts the cross-references exist — light docs-content test, catches link rot if a future doc edit removes a reference.Tier 3 — nit, not blocking
T3.1
.gitattributesdoesn't suppresspackage-lock.jsondiffs. Lockfile diffs are usually noisy when version bumps happen (PR #108'sset-version.mjswill mutate it). Addingpackage-lock.json -diffwould tell git to show "binary diff" instead of line-by-line, which keeps PR review surface clean. Not load-bearing — the lockfile diff is still meaningful sometimes — but cheaper to add now than after the first noisy version-bump PR. Worth a single follow-up commit if convenient.Cross-cutting note
This PR is the cleanest example of "Gemini explored broadly, Codex disciplined into reviewable slice" pattern across the three you sent. The fact that
CONTRIBUTING.mdcodifies the very rule that makes this slice possible ("One issue per PR. Keep review units small. Do not batch unrelated docs, scripts, and analyzer behavior together.") is recursive in a satisfying way.Verdict
approve_merge. Docs-only, scope-disciplined, partially closes the issue #17 substrate-gap I flagged twelve days ago. The next step toward fully closing #17 S5 ("nodecisions/directory, no ADRs") is still ahead — the CONTRIBUTING rules deserve eventual numbered ADRs so they survive conversation compaction the way fallow-py's do. Not in this PR's scope.—
claude(Opus 4.7), sister-project pass 🎸View command line instructions
Checkout
From your project repository, check out a new branch and test the changes.Merge
Merge the changes and update on Forgejo.Warning: The "Autodetect manual merge" setting is not enabled for this repository, you will have to mark this pull request as manually merged afterwards.