Uh oh!
There was an error while loading. Please reload this page.
refactor(tools): flatten the page and stop the Bash section from vanishing - #86
refactor(tools): flatten the page and stop the Bash section from vanishing#86Fl0p wants to merge 3 commits into
Conversation
…shing The tools table was wrapped in a Card titled "All Tools". DataTable already draws its own bordered, shadowed surface, so that was a box inside a box for a heading the page title already gave. The table now sits directly on the page, matching the Users list. The Bash breakdown rendered as null whenever it had no rows, so it disappeared with no trace and no explanation. It is now a plain section that always renders and states why it is empty. It is empty for a reason no fix can change: Claude Code's tool spans carry tool_name, tool_use_id and duration_ms only — which tool ran, not what it ran. Captured from a live session (enhanced-telemetry beta, and with user-prompt logging on) no span carries a command or tool_input attribute at any point. The endpoint stays for OTLP producers that do send one; the docs and changelog now say so instead of implying the breakdown works against Claude Code. Co-Authored-By: Daedalus <daedalus@agents.flopbut.local> Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Warning Review limit reached
Next review available in:40 minutes You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (3)
📝 WalkthroughWalkthroughThe Tools page now uses a flat layout and always displays the Bash commands section. It shows an explanatory empty state when command data is unavailable. Documentation describes the Claude Code telemetry limitation, and dashboard asset references use updated hashed filenames. ChangesTools page updates
Estimated code review effort: 2 (Simple) | ~10 minutes 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches📝 Generate docstrings
🧪 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 |
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 `@internal/dashboard/static/index.html`:
- Around line 8-9: The dashboard entrypoint in index.html references hashed Vite
assets that are absent from internal/dashboard/static, causing 404s. Either add
the corresponding built JavaScript and CSS assets under the static assets
directory, or update the script and stylesheet references to filenames that
exist there, while preserving a working dashboard load.
🪄 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: 6bb0d265-6de6-4de2-a9f5-ec5b0d40abb0
📒 Files selected for processing (5)
CHANGELOG.mddocs/operations/api-reference.mdfrontend/src/pages/Tools.module.cssfrontend/src/pages/Tools.tsxinternal/dashboard/static/index.html
Uh oh!
There was an error while loading. Please reload this page.
…lemetry The Bash section only branched on loading and row count, so a failed /bash-commands request fell through to the empty state and told the user Claude Code sends no command text — attributing an API failure to the producer's telemetry. Destructure SWR's error and render ErrorState. Co-Authored-By: Daedalus <daedalus@agents.flopbut.local> Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Daedalus <daedalus@agents.flopbut.local> Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fl0p
commented
Aug 9, 2026
Squash-merged locally as |
Follow-up to the board's feedback on the tools page rework: the nesting was
redundant, and the Bash breakdown was invisible with no way to tell why.
Flatten
DataTablealready draws its own bordered, shadowed surface, so wrapping it ina
Card title="All Tools"rendered a box inside a box — for a heading the pagetitle already gave. The table now sits directly on the page, matching the Users
list. The Bash block loses its card too and becomes a plain
<section>with aheading, so the page reads as one flat column.
Why the Bash breakdown was never visible
It rendered as
nullwhenever it had no rows, so it vanished entirely — noheading, no empty state, nothing to explain the absence. It now always renders
and states the reason.
The reason is not a bug. I captured a live Claude Code session against a local
OTLP receiver. Its Bash tool span is:
Which tool ran, not what it ran. Across the full session — every span type, with
CLAUDE_CODE_ENHANCED_TELEMETRY_BETA=1— nocommandand notool_inputattribute appears anywhere;
user_promptcomes through<REDACTED>. So thisbreakdown cannot be populated by Claude Code telemetry at all.
That matters because the 0.3.0 changelog reads as if the DuckDB filter-pushdown
fix made this work. The pushdown bug was real and the fix is correct, but it
could not on its own put rows in this table. The docs and changelog now say so.
The endpoint is kept for OTLP producers that do send
command— cotel's ingestis generic.
Open product question for the board: since this can never populate for a
Claude Code user, the alternative is deleting the section and the endpoint
outright rather than showing a permanently-empty block. I kept it (reversible,
and it answers the question in-product) but the removal is a one-liner if you'd
rather not carry dead surface.
Verified
npm run buildgreen;go build ./...andgo test ./internal/api/green(docker
golang:1.23-bookworm)spans (no
commandattribute — reproducing the board's view):/api/v1/tools?range=month&sort=avg_duration_ms&order=asc&page=1&limit=0🤖 Generated with Claude Code
Summary by CodeRabbit
New Features
Documentation
Style