fix(web): stop showing red x summaries for ordinary tool failures

[t3dotgg]

The red x shows up far too often as the summary of a set of tool calls or commands. A single nonzero exit at the end of a group paints the whole collapsed row red, which reads like something is broken when the agent usually just recovered and moved on.

Now the top level stays calm:

• Collapsed tool group summaries keep their normal muted icon. No red x, no red text.
• The "+N previous tool calls" fold always shows its chevron.
• The live activity row loses its red tint.

When you unfold the group, failed rows still show an x, but muted like every other icon instead of red. Screen readers keep the "tool call failed" hint everywhere it existed before.

Red treatment is reserved for severe failures: connection banners, runtime errors, and warnings keep their current styling. Mobile gets the matching change: the per-row failure glyph drops its red tint, while runtime error and warning rows stay rose.

	Before	After
Collapsed
Expanded

Follow-up after auditing real usage data (4,042 turns, ~5,300 ordinary tool failures):

• Orchestration failure activities (provider.*.failed, checkpoint.*.failed) keep the red row treatment, same as runtime.error. These are rare (5 in the audited data) and genuinely severe. This closes the gap Bugbot flagged.
• Runtime warnings with no displayable text (background_tasks_changed and similar adapter passthrough, 300+ rows in the audited data) no longer render on web or mobile. They carried nothing a user can act on but drew an x-shaped warning icon, rose tinted on mobile. Warnings with real content, such as provider stream retry notices, still render.

Changes made by Claude Fable 5 in Claude Code.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: e2f745fa-feff-4f4e-a9b5-71b855aa7ac9

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.3 KiB	13.3 KiB	−12 B (−0.1%)	15.1 KiB	✅
Codex	Thread snapshot wire	6.9 KiB	6.9 KiB	+7 B (+0.1%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.4 KiB	6.4 KiB	−19 B (−0.3%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	55.6 KiB	55.6 KiB	−44 B (−0.1%)	66.4 KiB	✅
Codex	Live turn messages	11	10	−1 (−9.1%)	21	✅
Claude	Total thread wire	13.4 KiB	13.3 KiB	−70 B (−0.5%)	15.1 KiB	✅
Claude	Thread snapshot wire	6.9 KiB	6.9 KiB	+8 B (+0.1%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.5 KiB	6.4 KiB	−78 B (−1.2%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	56.4 KiB	56.4 KiB	−88 B (−0.2%)	66.4 KiB	✅
Claude	Live turn messages	11	9	−2 (−18.2%)	21	✅

Baseline: a6797b3 · PR result: dd46538 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 109.4 KiB
• Claude decoded thread snapshot: 110.1 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit b9d657e. Configure here.}

[macroscopeapp]

Approvability

Verdict: Approved at dd46538

Macroscope's review found this PR approvable — This is a bounded UI bug fix that tones down ordinary tool-failure indicators while preserving accessibility and severe-failure signaling. The related filtering only removes explicitly non-displayable warning noise, with focused web and mobile tests covering the changed behavior.

^{You can add or adjust custom eligibility rules. Learn more.}