feat(agents): request summarized Claude thinking and label empty rows

[wpfleger96]

Claude Code agents on Opus showed Thinking rows that opened to nothing. Opus returns no readable thinking unless Claude Code is asked for summaries, and Buzz never asked. Sonnet returns readable thinking either way.

Request summaries from Claude Code. session_new_full now adds _meta.claudeCode.options.extraArgs: {"thinking-display": "summarized"} for the Claude adapter. claude-agent-acp spreads extraArgs into the CLI's arguments; 0.36.1, 0.60.0 and 0.66.0 all read the same shape. The value is merged into the existing _meta, so systemPrompt and sessionTitle are kept. session_new_full is the only place Buzz sends session/new, so agent sessions and model discovery both get it.

Only for a CLI that accepts the flag. --thinking-display first shipped in Claude Code 2.1.94, and older CLIs exit on unknown flags, which would fail every session start. When the adapter is spawned, Buzz runs $CLAUDE_CODE_EXECUTABLE --version once with a 5-second limit, reading at most 256 bytes of output and killing its whole process group on timeout, and sends the flag only on 2.1.94 or newer. The flag is left off, and sessions start exactly as before, when:

• the CLI is older than 2.1.94;
• CLAUDE_CODE_EXECUTABLE is unset (the adapter then runs its own bundled CLI, which Buzz can't inspect) or empty;
• the binary hangs, fails, or prints anything other than a clean major.minor.patch version;
• BUZZ_ACP_LAUNCH_PREFIX is set, since the wrapper can change which CLI the adapter actually runs.

The result is logged once at debug level.

Label rows that still have no text. A Thinking row whose text is empty or only whitespace keeps its header and shows a muted "No readable reasoning tokens", with nothing to expand. This covers models or CLIs that still return no readable thinking. If text arrives later in the same row, it renders as an expandable section as before. ThoughtActivity is the only component that draws Thinking rows.

[github-actions]

🔐 Codex Security Review

Note: This is an automated, security-focused review generated by Codex.
Use it as a supplement to human review; false positives are possible.

Scope

• Exact PR diff: 133fb9869228dab4c00cb2b9e259bf834a55dbf7...c27444a35fb52be53588054bfaa7067ac5ba8308
• Model: gpt-5.6-sol

💡 Click "edited" above to see earlier reviews for this PR.

---

Review Summary

Overall Risk: HIGH

The Claude version probe bypasses the adapter subprocess environment sanitization and can expose an ambient Nostr signing key.

Findings

[HIGH] Claude version probe inherits a private key removed from the adapter environment

• Category: Cryptography
• Location: crates/buzz-acp/src/acp.rs:2457 (source)
• Description: The adapter command explicitly removes NOSTR_PRIVATE_KEY before spawning, and existing integration coverage treats that as a required secret boundary. The new capability check creates an independent Command for CLAUDE_CODE_EXECUTABLE. That command inherits the harness process environment rather than the sanitized adapter environment, so an ambient NOSTR_PRIVATE_KEY is available during execution of --version. This occurs before the protected adapter is spawned and can expose a distinct operator or owner key to the configured Claude executable.
• Impact: A compromised or substituted Claude executable can steal the ambient Nostr private key and use it to sign events, impersonate that identity, or access resources authorized to it.
• Recommendation: Run the probe with the same effective sanitized environment as the adapter command, including all env removals, or use env_clear and copy only the minimal variables required by the version executable. Add a regression test that places a canary in NOSTR_PRIVATE_KEY and verifies the version probe cannot observe it.

Notes

• Review used read-only source and history inspection as required; repository tests and binaries were not executed.

---

_{Generated by Codex Security Review |
Requested by: @wpfleger96 |
Workflow run}

[wesbillman]

Carl, an automated reviewer, commenting via Wes’s GitHub account.

No blocking findings. Reviewed head 5e6b88a78346a704bec4927df3c06bde0e25b08d against base 448407a972ca9da0c1e13d49ee2c2170821be8a2, including independent UI and Claude-adapter review. One optional subprocess-hardening note is inline. This is a comment, not approval.

Validation: hosted Rust checks passed; Desktop passed 6,779 ordinary JS tests and 133 jsdom tests, including all three new renderer cases. CI tested merge cfa3b4143575768425644682184f188344bf16d6; its entire Desktop tree and all five changed files match the reviewed head. Confirmed the extraArgs contract against upstream adapter versions 0.36.1, 0.60.0 and 0.66.0.

Remaining gates: smoke shard 2 failed the unchanged empty-edit-delete.spec.ts:95 message-edit test on every attempt; I found no connection to this PR’s changed paths, but CI is not green. Security review is still running. Real authenticated Claude/Opus output and human acceptance remain unverified. Resolve or explicitly disposition the CI failure and verify the live thinking flow before treating this as ready to merge.