Review: 44a884e#2
Conversation
|
/oc why is this not showing up in linear review? |
|
I can't tell from this repo — Linear's review queue is configured on the Linear side (GitHub integration settings, label/repo filters, branch rules). A few things to check there:
Nothing in the repo ( |
|
/oc will adding me as a reviewer while making pr inside the action fix this? read |
|
No. Linear's GitHub integration typically keys off the PR author, not reviewers — and this PR is authored by Likely fixes, in order of probability:
If the goal is just "make sure a human sees it before merge," assigning yourself is fine — but it won't make Linear pick it up. |
|
/oc inside this page i clicked on request a review for myself and i it got to show up in linear how do i do this with acitons. |
|
/oc inside this page i clicked on request a review for myself and i it got to show up in linear how do i do this with acitons. |
|
/oc inside this page i clicked on request a review for myself and i it got to show up in linear how do i do this with acitons. |
|
The trigger is the review-request event, not the author. When you click "Request a review" on the PR page, Linear's GitHub integration sees the To make the action do the same, it needs to add you as a reviewer after the PR is created. Either: gh pr edit "$PR_URL" --add-reviewer samarth-naor via the API: curl -X POST \
-H "Authorization: Bearer $GITHUB_TOKEN" \
-H "Accept: application/vnd.github+json" \
https://api.github.com/repos/samarth-na/dotfiles/pulls/<num>/requested_reviewers \
-d '{"reviewers":["samarth-na"]}'Add that step right after the Heads up: there's no |
|
Add yourself as a requested reviewer right after the PR is created. Linear's GitHub integration syncs on the gh pr edit "$PR_URL" --add-reviewer samarth-naor via the API: curl -X POST \
-H "Authorization: Bearer $GITHUB_TOKEN" \
-H "Accept: application/vnd.github+json" \
https://api.github.com/repos/samarth-na/dotfiles/pulls/<num>/requested_reviewers \
-d '{"reviewers":["samarth-na"]}'Drop it as a step right after |
| @@ -1,21 +1,25 @@ | |||
| # AGENTS.md | |||
|
|
|||
| ## Scope | |||
There was a problem hiding this comment.
/oc is this optimal bash code
|
No. Key issues:
Quick hardening: #!/usr/bin/env bash
set -euo pipefail
[[ -n "${TMUX:-}" ]] || { echo "Error: Not in a tmux session" >&2; exit 1; }
command -v fzf-tmux >/dev/null || { echo "Error: fzf-tmux not found" >&2; exit 1; }
target=$(tmux list-panes -a -F '#{session_name}:#{window_index}.#{pane_index}.#{window_name}-#{pane_current_path}' \
| fzf-tmux -p | cut -d "." -f 1)
[[ -n "$target" ]] && tmux switch-client -t "$target" |





Code Review
Automated review of the last commit on
dev.Review Output
Errors from
switch-clientare redirected to/dev/nulland masked by|| true. The fish version at least prints a message to stderr (or echo "" >&2). This makes debugging failures (e.g., malformed target from a session name containing dots) impossible.3. No
fzf-tmuxDependency CheckIf
fzf-tmuxis not installed (or tmux popup mode is unavailable), the pipeline fails with a cryptic error from the shell, not a helpful message.4. Missing AGENTS.md Update
The new script is not documented in
AGENTS.mdunder Layout or Gotchas, so future readers won't know it exists or that it's a bash equivalent of the fish function.What's Fine
cut -d "." -f 1parsing works for typical session names; the same logic exists in the fish version, so not a regression.Bottom line: Remove the orphaned bash script unless there's a concrete plan to wire it up from bash contexts, or wire it up and update AGENTS.md.