Repository navigation
docs: weekly update for PRs merged 2026-07-29 to 2026-08-05 - #6206
Merged
Merged
Conversation
- tasks.mdx: Tasks and Pull requests are now separate sidebar destinations with independent persisted filters; GitHub Issues as a Tasks source (#5975) - workspaces.mdx: prompt history popover (#6083), typed-name-seeds-branch behavior (#6129), right-click import of untracked worktrees (#6097), bulk workspace actions section (#6021) - automations.mdx: inline Retry and Retry all for failed runs (#6139) - keyboard-shortcuts.mdx: Open Command Palette ⌘⇧K and Check Resources ⌘⇧U (#6037) - setup-teardown-scripts.mdx: experimental wait-for-setup-before-agents setting (#5925)
Contributor
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (3)
🚧 Files skipped from review as they are similar to previous changes (3)
📝 WalkthroughWalkthroughThe documentation covers automation retries, keyboard shortcuts, setup-gated agent startup, task and pull-request views, workspace actions, and agent-driven Superset workflows. ChangesProduct documentation updates
Estimated code review effort: 2 (Simple) | ~10 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Contributor
Greptile SummaryThe PR updates five documentation pages for recently added automation, workspace, task, pull-request, and keyboard-shortcut behavior.
Confidence Score: 4/5The PR appears safe to merge after minor documentation corrections for branch-name normalization and the Mine-only Retry-all scope. The documented features generally match their implementations, but two statements can lead users to expect an exact branch name or a Retry-all control in a scope where it is unavailable. Files Needing Attention: apps/docs/content/docs/workspaces.mdx, apps/docs/content/docs/automations.mdx
|
| Filename | Overview |
|---|---|
| apps/docs/content/docs/automations.mdx | Documents failed-run retry controls, but does not state that Retry all is available only in the Mine scope. |
| apps/docs/content/docs/keyboard-shortcuts.mdx | Adds two shortcut entries that match the registered macOS bindings and command labels. |
| apps/docs/content/docs/setup-teardown-scripts.mdx | Documents the experimental setup-before-agent sequencing behavior consistently with command chaining. |
| apps/docs/content/docs/tasks.mdx | Updates navigation, source, and filter documentation consistently with the separate Tasks and Pull Requests routes. |
| apps/docs/content/docs/workspaces.mdx | Adds several workspace features, but inaccurately describes normalized workspace names as being used directly for branch names. |
Prompt To Fix All With AI
### Issue 1
apps/docs/content/docs/workspaces.mdx:22
**Branch names are normalized**
Manually entered workspace names are normalized before becoming branch names: letters are lowercased, spaces become hyphens, punctuation is removed, and values are truncated to 25 characters. Saying the name is used “directly” causes users to look for an exact branch name that was never created.
```suggestion
If you type a workspace name yourself, Superset derives a branch name from it by normalizing the text. Otherwise Superset generates a name from your prompt.
```
### Issue 2
apps/docs/content/docs/automations.mdx:65
**Retry all is Mine-only**
The **Retry all** button is rendered only in the **Mine** scope, even when the user owns failed automations visible from another scope. Without that qualification, users viewing **Team** satisfy the documented condition but cannot find the advertised control.
```suggestion
In the Automations list, hover an automation whose last run failed (`skipped (offline)` or `dispatch failed`) to reveal an inline **Retry** button that re-dispatches it. In the **Mine** view, a **Retry all** button appears in the header when any of your automations have failed last runs, re-dispatching all of them at once. Both disappear automatically once nothing is failed.
```
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.Reviews (1): Last reviewed commit: "docs: weekly update for PRs merged 2026-..." | Re-trigger Greptile
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
| In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. | ||
| In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. The prompt box also keeps a searchable history of your previous prompts — click the history icon to reuse one. | ||
|
|
||
| If you type a workspace name yourself, it's used directly as the branch name. Otherwise Superset generates a name from your prompt. |
Contributor
There was a problem hiding this comment.
Manually entered workspace names are normalized before becoming branch names: letters are lowercased, spaces become hyphens, punctuation is removed, and values are truncated to 25 characters. Saying the name is used “directly” causes users to look for an exact branch name that was never created.
Suggested change
| If you type a workspace name yourself, it's used directly as the branch name. Otherwise Superset generates a name from your prompt. | |
| If you type a workspace name yourself, Superset derives a branch name from it by normalizing the text. Otherwise Superset generates a name from your prompt. |
Prompt To Fix With AI
This is a comment left during a code review.
Path: apps/docs/content/docs/workspaces.mdx
Line: 22
Comment:
**Branch names are normalized**
Manually entered workspace names are normalized before becoming branch names: letters are lowercased, spaces become hyphens, punctuation is removed, and values are truncated to 25 characters. Saying the name is used “directly” causes users to look for an exact branch name that was never created.
```suggestion
If you type a workspace name yourself, Superset derives a branch name from it by normalizing the text. Otherwise Superset generates a name from your prompt.
```
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!
|
|
||
| ## Retry Failed Runs | ||
|
|
||
| In the Automations list, hover an automation whose last run failed (`skipped (offline)` or `dispatch failed`) to reveal an inline **Retry** button that re-dispatches it. A **Retry all** button appears in the header when any of your automations have failed last runs, re-dispatching all of them at once. Both disappear automatically once nothing is failed. |
Contributor
There was a problem hiding this comment.
The Retry all button is rendered only in the Mine scope, even when the user owns failed automations visible from another scope. Without that qualification, users viewing Team satisfy the documented condition but cannot find the advertised control.
Suggested change
| In the Automations list, hover an automation whose last run failed (`skipped (offline)` or `dispatch failed`) to reveal an inline **Retry** button that re-dispatches it. A **Retry all** button appears in the header when any of your automations have failed last runs, re-dispatching all of them at once. Both disappear automatically once nothing is failed. | |
| In the Automations list, hover an automation whose last run failed (`skipped (offline)` or `dispatch failed`) to reveal an inline **Retry** button that re-dispatches it. In the **Mine** view, a **Retry all** button appears in the header when any of your automations have failed last runs, re-dispatching all of them at once. Both disappear automatically once nothing is failed. |
Prompt To Fix With AI
This is a comment left during a code review.
Path: apps/docs/content/docs/automations.mdx
Line: 65
Comment:
**Retry all is Mine-only**
The **Retry all** button is rendered only in the **Mine** scope, even when the user owns failed automations visible from another scope. Without that qualification, users viewing **Team** satisfy the documented condition but cannot find the advertised control.
```suggestion
In the Automations list, hover an automation whose last run failed (`skipped (offline)` or `dispatch failed`) to reveal an inline **Retry** button that re-dispatches it. In the **Mine** view, a **Retry all** button appears in the header when any of your automations have failed last runs, re-dispatching all of them at once. Both disappear automatically once nothing is failed.
```
---
For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@apps/docs/content/docs/automations.mdx`:
- Around line 63-66: Update the retry-actions description in the Automations
documentation to say both buttons disappear “when no automation has a failed
latest run” instead of “once nothing is failed,” while preserving the
surrounding behavior and terminology.
In `@apps/docs/content/docs/tasks.mdx`:
- Line 12: Clarify the source scope in the Tasks view description: distinguish
the behavior for connected Linear issues from Superset-native tasks and GitHub
Issues, and state whether the workspace-wide listing applies only to Linear or
to every source.
🪄 Autofix
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 703ef9a6-0d6c-42ec-98c6-51fb57dc3b15
📒 Files selected for processing (5)
apps/docs/content/docs/automations.mdxapps/docs/content/docs/keyboard-shortcuts.mdxapps/docs/content/docs/setup-teardown-scripts.mdxapps/docs/content/docs/tasks.mdxapps/docs/content/docs/workspaces.mdx
Contributor
🧹 Preview Cleanup CompleteThe following preview resources have been cleaned up:
Thank you for your contribution! 🎉 |
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
1 issue found across 5 files
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="apps/docs/content/docs/workspaces.mdx">
<violation number="1" location="apps/docs/content/docs/workspaces.mdx:22">
P2: This user-facing sentence says typing a workspace name uses it directly as the branch name, but in the source the workspace name and branch name are separate fields. The workspace-name input's onChange only updates the workspace name; the branch name is only set from the explicit branch-name input or from AI generation from the prompt. So a typed workspace name is not what becomes the branch name. Recommend rewording to reflect that the branch name is a separate, independently editable field.</violation>
</file>
Architecture diagram
sequenceDiagram
participant User as User (Dev)
participant UI as Desktop App UI
participant Sidebar as Sidebar Nav
participant Hotkeys as Hotkey Registry
participant Workspace as Workspace Manager
participant Automation as Automation Engine
participant Setup as Setup Script Runner
participant GitHub as GitHub API
participant Linear as Linear API
Note over User,Linear: Tasks & Pull Requests Split
User->>Sidebar: Click "Tasks" destination
Sidebar->>Sidebar: Load independently persisted filters
Sidebar->>Linear: Fetch issues with filter params
Sidebar->>GitHub: Fetch GitHub Issues as source
GitHub-->>Sidebar: Issues list
Linear-->>Sidebar: Issues list
Sidebar-->>User: Display tasks
User->>Sidebar: Click "Pull requests" destination
Sidebar->>Sidebar: Load repository/state/search filters
Sidebar->>GitHub: Fetch PRs with filter params
GitHub-->>Sidebar: PR list
Sidebar-->>User: Display PRs
Note over User,Setup: Workspace Creation Flow
User->>UI: Press ⌘N or click "New Workspace"
UI->>UI: Show prompt history popover (clickable)
User->>UI: Select/type workspace name
UI->>Workspace: Use typed name as branch name directly
Workspace-->>UI: Workspace created
alt Import untracked worktrees
User->>UI: Right-click project > Import untracked worktrees
UI->>UI: Show confirmation dialog with branches/paths
alt Run setup script checked
UI->>Setup: Execute setup script on imported worktree
Setup-->>UI: Setup complete
else Run setup script unchecked (default)
Note over UI,Setup: No setup run (already set up)
end
UI-->>User: Worktree imported
end
alt Bulk Actions on Workspaces
User->>UI: ⌘-click or Shift-click select multiple workspaces
UI->>UI: Show toolbar with Move/Ungroup/Delete
User->>UI: Click Delete
UI->>UI: Preview dirty/unpushed changes
UI-->>User: Confirm deletion
end
Note over User,Automation: Retry Failed Runs
User->>UI: View automations list
UI->>Automation: Fetch last run status for each automation
Automation-->>UI: Return run statuses
alt Automation has failed last run
UI->>UI: Show inline Retry button per-row
User->>UI: Click inline Retry
UI->>Automation: Re-dispatch automation
Automation-->>UI: Dispatch queued
else Multiple failures exist
UI->>UI: Show Retry all button in header
User->>UI: Click Retry all
UI->>Automation: Re-dispatch all failed automations
Automation-->>UI: All dispatched
end
Note over User,Setup: Experimental Agent Start Setting
User->>UI: Settings > Experimental > Enable "Wait for workspace setup before starting agents"
UI->>Setup: Configure agent launch behavior
User->>Workspace: Create workspace (agent included)
alt Setting enabled
Workspace->>Setup: Run setup commands first
Setup-->>Workspace: Setup succeeded
Workspace->>Workspace: Start agent in same terminal after setup
else Setting disabled (default)
Workspace->>Workspace: Start agent immediately, parallel with setup
end
Note over User,Hotkeys: New Keyboard Shortcuts
User->>Hotkeys: Press ⌘⇧K (Open Command Palette)
Hotkeys->>UI: Open command palette overlay
UI-->>User: Command palette displayed
User->>Hotkeys: Press ⌘⇧U (Check Resources)
Hotkeys->>UI: Show resource check dialog
UI-->>User: Resources status shown
Reply with feedback, questions, or to request a fix.
Re-trigger cubic
| In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. | ||
| In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. The prompt box also keeps a searchable history of your previous prompts — click the history icon to reuse one. | ||
|
|
||
| If you type a workspace name yourself, it's used directly as the branch name. Otherwise Superset generates a name from your prompt. |
Contributor
There was a problem hiding this comment.
P2: This user-facing sentence says typing a workspace name uses it directly as the branch name, but in the source the workspace name and branch name are separate fields. The workspace-name input's onChange only updates the workspace name; the branch name is only set from the explicit branch-name input or from AI generation from the prompt. So a typed workspace name is not what becomes the branch name. Recommend rewording to reflect that the branch name is a separate, independently editable field.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At apps/docs/content/docs/workspaces.mdx, line 22:
<comment>This user-facing sentence says typing a workspace name uses it directly as the branch name, but in the source the workspace name and branch name are separate fields. The workspace-name input's onChange only updates the workspace name; the branch name is only set from the explicit branch-name input or from AI generation from the prompt. So a typed workspace name is not what becomes the branch name. Recommend rewording to reflect that the branch name is a separate, independently editable field.</comment>
<file context>
@@ -17,7 +17,9 @@ Press **⌘N** or click "New Workspace". Create from:
-In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly.
+In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. The prompt box also keeps a searchable history of your previous prompts — click the history icon to reuse one.
+
+If you type a workspace name yourself, it's used directly as the branch name. Otherwise Superset generates a name from your prompt.
Each workspace gets:
</file context>
Suggested change
| If you type a workspace name yourself, it's used directly as the branch name. Otherwise Superset generates a name from your prompt. | |
| The branch name is a separate field you can type directly; otherwise Superset generates one from your prompt. |
…set:* skills Covers the workflow from #6181: standup, automate, doctor, setup, feedback, and orchestrate driven from the agent you're already in.
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@apps/docs/content/docs/recipes/agent-driven-superset.mdx`:
- Line 16: Update Step 1 in the agent-driven Superset recipe to distinguish
skill namespaces by agent: retain the `superset:` examples for Claude Code, and
identify Codex and other skills-capable agents as using the `superset-`
namespace. Ensure the expected skill names are qualified accordingly so users
can verify installations without assuming the wrong names.
🪄 Autofix
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: fe8cb1c5-0c11-4e65-8296-cbeb0354af3e
📒 Files selected for processing (2)
apps/docs/content/docs/meta.jsonapps/docs/content/docs/recipes/agent-driven-superset.mdx
|
|
||
| ## Steps | ||
|
|
||
| 1. Confirm the skills are there: ask any agent launched through Superset to "list your superset skills". You should see `superset:standup`, `superset:automate`, `superset:doctor`, and friends. (No desktop app on this machine? `npx skills add superset-sh/skills`.) |
Contributor
There was a problem hiding this comment.
🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -euo pipefail
curl -fsSL https://docs.superset.sh/skills | rg -n 'Claude Code|Codex|superset:|superset-'Repository: superset-sh/superset
Length of output: 50376
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "## target file context"
sed -n '1,80p' apps/docs/content/docs/recipes/agent-driven-superset.mdx
echo
echo "## skill namespace references in docs"
rg -n "Claude Code|Codex|Kim|i|Vibe|Grok|superset:|superset-" apps/docs/content docs 2>/dev/null || true
echo
echo "## external /skills focused relevant lines"
python3 - <<'PY'
import urllib.request, re
from html.parser import HTMLParser
url = "https://docs.superset.sh/skills"
with urllib.request.urlopen(url, timeout=20) as r:
html = r.read().decode("utf-8", errors="replace")
text = "".join(HTMLParser().parse(html).data.splitlines())
# Extract visible prose around namespace statements.
for needle in ["Claude Code", "Codex", "superset:", "superset-", "slash commands", "plugins"]:
idx = text.find(needle)
if idx != -1:
start = max(0, idx - 400)
end = min(len(text), idx + 1200)
# Strip script/style text lightly before printing
snippet = re.sub(r"<(?:script|style)[^>]*>.*?</(?:script|style)>", " ", text[start:end], flags=re.S)
print("### around", needle)
print(snippet.replace('\n', ' '))
print()
PYRepository: superset-sh/superset
Length of output: 50379
🏁 Script executed:
#!/bin/bash
set -euo pipefail
echo "## target file"
sed -n '1,70p' apps/docs/content/docs/recipes/agent-driven-superset.mdx
echo
echo "## skills docs focused match"
sed -n '1,40p' apps/docs/content/docs/skills.mdx | cat -n
echo
echo "## focused namespace references in docs"
rg -n "Claude Code|Codex|skills|superset:|superset-" apps/docs/content/docs/recipes/agent-driven-superset.mdx apps/docs/content/docs/skills.mdx apps/docs/content/docs/mcp-server.mdx --iglob '*.mdx' --context 2Repository: superset-sh/superset
Length of output: 15329
Use the correct skill namespace for each agent.
Step 1 says users “should see” superset:standup, superset:automate, and superset:doctor, but /skills documents that Claude Code uses superset: while Codex and other skills-capable agents use superset-. Qualify the expected skill names by agent so non-Claude users do not treat installed skills as missing.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@apps/docs/content/docs/recipes/agent-driven-superset.mdx` at line 16, Update
Step 1 in the agent-driven Superset recipe to distinguish skill namespaces by
agent: retain the `superset:` examples for Claude Code, and identify Codex and
other skills-capable agents as using the `superset-` namespace. Ensure the
expected skill names are qualified accordingly so users can verify installations
without assuming the wrong names.
… defer to skills page
Contributor
There was a problem hiding this comment.
1 issue found across 2 files (changes from recent commits).
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="apps/docs/content/docs/recipes/agent-driven-superset.mdx">
<violation number="1" location="apps/docs/content/docs/recipes/agent-driven-superset.mdx:16">
P3: The listed skill names (`superset:standup`, `superset:automate`, `superset:doctor`) use the `superset:` namespace, which per the `/skills` docs applies to Claude Code only — Codex and other skills-capable agents use `superset-` instead. Consider qualifying the expected skill names by agent so non-Claude users don't think the skills failed to install.</violation>
</file>
Tip: Review your code locally with the cubic CLI to iterate faster.
Re-trigger cubic
|
|
||
| ## Steps | ||
|
|
||
| 1. Confirm the skills are there: ask any agent launched through Superset to "list your superset skills". You should see `superset:standup`, `superset:automate`, `superset:doctor`, and friends. (No desktop app on this machine? `npx skills add superset-sh/skills`.) |
Contributor
There was a problem hiding this comment.
P3: The listed skill names (superset:standup, superset:automate, superset:doctor) use the superset: namespace, which per the /skills docs applies to Claude Code only — Codex and other skills-capable agents use superset- instead. Consider qualifying the expected skill names by agent so non-Claude users don't think the skills failed to install.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At apps/docs/content/docs/recipes/agent-driven-superset.mdx, line 16:
<comment>The listed skill names (`superset:standup`, `superset:automate`, `superset:doctor`) use the `superset:` namespace, which per the `/skills` docs applies to Claude Code only — Codex and other skills-capable agents use `superset-` instead. Consider qualifying the expected skill names by agent so non-Claude users don't think the skills failed to install.</comment>
<file context>
@@ -0,0 +1,52 @@
+
+## Steps
+
+1. Confirm the skills are there: ask any agent launched through Superset to "list your superset skills". You should see `superset:standup`, `superset:automate`, `superset:doctor`, and friends. (No desktop app on this machine? `npx skills add superset-sh/skills`.)
+2. Start your day in any agent session instead of clicking through workspaces:
+
</file context>
…on the automate arc
Contributor
There was a problem hiding this comment.
1 issue found across 3 files (changes from recent commits).
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="apps/docs/content/docs/workspaces.mdx">
<violation number="1" location="apps/docs/content/docs/workspaces.mdx:22">
P3: The docs claim a typed workspace name is "slugified to kebab-case", but the create flow calls sanitizeBranchNameWithMaxLength with preserveFirstSegmentCase: true and prepends an author/prefix, so the first segment keeps its original casing and the result isn't strictly kebab-case. Consider wording like "cleaned into a valid branch name" or confirming the casing behavior against create.ts before shipping this phrasing.</violation>
</file>
Tip: Review your code locally with the cubic CLI to iterate faster.
Re-trigger cubic
| In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. | ||
| In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. The prompt box also keeps a searchable history of your previous prompts — click the history icon to reuse one. | ||
|
|
||
| If you type a workspace name yourself, the branch name is derived from it (slugified to kebab-case). Otherwise Superset generates a name from your prompt. |
Contributor
There was a problem hiding this comment.
P3: The docs claim a typed workspace name is "slugified to kebab-case", but the create flow calls sanitizeBranchNameWithMaxLength with preserveFirstSegmentCase: true and prepends an author/prefix, so the first segment keeps its original casing and the result isn't strictly kebab-case. Consider wording like "cleaned into a valid branch name" or confirming the casing behavior against create.ts before shipping this phrasing.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At apps/docs/content/docs/workspaces.mdx, line 22:
<comment>The docs claim a typed workspace name is "slugified to kebab-case", but the create flow calls sanitizeBranchNameWithMaxLength with preserveFirstSegmentCase: true and prepends an author/prefix, so the first segment keeps its original casing and the result isn't strictly kebab-case. Consider wording like "cleaned into a valid branch name" or confirming the casing behavior against create.ts before shipping this phrasing.</comment>
<file context>
@@ -19,7 +19,7 @@ Press **⌘N** or click "New Workspace". Create from:
In the prompt-first workspace flow, **Link pull request** lets you search by title or PR number, or paste the full PR URL to match it directly. The prompt box also keeps a searchable history of your previous prompts — click the history icon to reuse one.
-If you type a workspace name yourself, it's used directly as the branch name. Otherwise Superset generates a name from your prompt.
+If you type a workspace name yourself, the branch name is derived from it (slugified to kebab-case). Otherwise Superset generates a name from your prompt.
Each workspace gets:
</file context>
Suggested change
| If you type a workspace name yourself, the branch name is derived from it (slugified to kebab-case). Otherwise Superset generates a name from your prompt. | |
| If you type a workspace name yourself, it's cleaned into a valid branch name (spaces and special chars replaced, length-limited). Otherwise Superset generates a name from your prompt. |
This was referenced Aug 12, 2026
This branch was previously deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
What changed, by driving PR
recipes/agent-driven-superset.mdx(new) — "Drive Superset from Your Agent": fan out work withsuperset:orchestrateand schedule recurring chores withsuperset:automate, from the agent you're already talking to; links the Skills page for the full list. Registered inmeta.jsonunder Recipes (feat(desktop,cli,trpc): ship superset:* plugin skills to all agents and add private feedback pipeline #6181)tasks.mdx— Tasks and Pull requests are now separate sidebar destinations with independently persisted filters; the dedicated PR view has repository/state/search filters; GitHub Issues noted as a Tasks source (feat(desktop): separate tasks and pull requests in sidebar #5975)workspaces.mdx— prompt history popover in the new-workspace flow (feat(desktop): persist prompt history and add search popover to new-workspace views #6083); typed workspace name is used directly as the branch name (fix(host-service): seed branch from typed workspace name, rename to AI names before terminals start #6129); right-click Import untracked worktrees with Run-setup-script checkbox (feat(desktop,host-service): add right-click import of untracked worktrees #6097); new Bulk Actions section for multi-select Move/Ungroup/Delete (feat(desktop): add bulk workspace actions to sidebar #6021)automations.mdx— new Retry Failed Runs section: inline per-row Retry and header Retry-all for failed last runs (feat(desktop): add retry buttons for failed automations #6139)keyboard-shortcuts.mdx— added Open Command Palette ⌘⇧K and Check Resources ⌘⇧U, verified againstapps/desktop/src/renderer/hotkeys/registry.ts(feat(desktop): move resource monitor into the command palette #6037)setup-teardown-scripts.mdx— documented the experimental Wait for workspace setup before starting agents setting (feat(desktop,host-service): gate agent start on workspace setup #5925)Deliberately skipped: self-documenting PRs (#6181's own skills/orchestration page updates, #6176), docs/marketing PRs, onboarding and platform changes already covered by #6175/#6180, transient v1→v2 migration UX, bug fixes restoring expected behavior, and internal/perf/polish work.
Testing
bun run --cwd apps/docs typecheck(compiles all MDX) — clean after each commitbun run lint:fixthenbun run lint— exit 0apps/docs— none remainSummary by CodeRabbit