fix(cli): split newlines before wrapping in approval/sudo/clarify panels (#72580)

[JonthanaHanh]

Summary

Fixes #72580.

_wrap_panel_text() (3 copies: approval panel, sudo panel, clarify panel) passes multi-line commands directly to textwrap.wrap(), which treats newlines as whitespace. A heredoc command gets collapsed into a few long lines with literal \n characters, making it unreadable.

Changes

• cli.py: All 3 copies of _wrap_panel_text now split text on newlines first, then wrap each segment individually. This preserves the original line structure.

Testing

• py_compile: OK
• ruff check: All checks passed

[teknium1]

Thanks for identifying a real rendering defect: current main still sends the full approval command through textwrap.wrap() at cli.py:13070, and renders that result at cli.py:13119.

Problems

• The PR does not cover all active wrappers it describes. The clarify wrapper remains unsplit at cli.py:16100 and is used for the displayed question at cli.py:16278.
• The first hunk is in _get_slash_confirm_display_fragments (cli.py:8648), not the sudo UI; its detail is already split before wrapping at cli.py:8685 and cli.py:8695. The sudo display is the fixed password panel at cli.py:16350.
• No regression test is included. tests/cli/test_cli_approval_ui.py:158 exercises a long single-line command, not newline preservation.

Suggested changes

• Apply the same split-before-wrap behavior to the clarify wrapper while preserving its distinct textwrap options.
• Add rendering tests for multiline approval and clarify text, plus a one-line control case.

Automated hermes-sweeper review.

[GottZ]

This was generated by AI during triage.

Summary

Four PRs address #72580 with split-before-wrap changes: #72601 fixes only slash confirmation and approval, #72614 also fixes clarify and adds approval tests, #73584 changes only approval but mishandles blank lines and bundles unrelated work, and #75207 centralizes all three wrappers while preserving clarify-specific semantics and adding the broadest focused tests.

Related pull requests

• #72601 partial — (+28/-16) — duplicate of #75207: the diff fixes slash confirmation and approval but leaves clarify unchanged and adds no regression tests. Despite the keep_open review on #72601, #75207 covers all three wrapper paths, preserves blank lines, and tests the shared behavior.
• #72614 best fix — (+78/-24) — duplicate of #75207: the diff correctly fixes slash confirmation, approval, and clarify while adding approval heredoc and one-line regressions, but retains three copies and lacks multiline clarify coverage. Despite the recorded best-fix and keep_open verdicts on #72614, #75207 implements the same correction centrally, retains clarify's distinct wrapping options, and tests clarify delegation and empty lines.
• #73584 duplicate — (+139/-149) — duplicate of #75207: its approval-only hunk renders an empty source line as visible '', has no regression tests, and is mixed with unrelated tool-hook and Himalaya documentation changes. Despite the keep_open review on #73584, the contributor-identified blank-line defect remains in its diff, while #75207 uses "" and tests blank-line preservation.
• #75207 best fix — (+152/-29) — keep open with a salvage path: consistent with its recorded best-fix and keep_open verdicts, the diff centralizes newline-preserving wrapping for slash confirmation and approval, delegates clarify with its distinct prose options, preserves empty lines, and adds focused regressions. The remaining salvage step is a true clarify-panel rendering-path test, because the amended tests call the delegation helper directly rather than exercising the panel binding.

Duplicates

#72601, #72614, and #73584 substantially duplicate the split-before-wrap correction in #75207; #75207 subsumes their relevant #72580 changes without #73584's blank-line defect or unrelated files.

Suggested consolidation

Keep #75207 open with a salvage path: retain its shared newline-preserving helpers and focused tests, then add a clarify-panel rendering regression that exercises the actual panel binding. Close #72601 and #73584 as duplicates of #75207; despite #72614's recorded best-fix and keep_open verdicts, close it as a duplicate because #75207 contains the same complete wrapper fix with centralized logic and broader empty-line and clarify-delegation coverage. Keep #73584's unrelated tool-hook and Himalaya changes independently reviewable.

Complex graph

flowchart LR
    classDef open fill:#dbeafe,stroke:#1d4ed8,color:#1e3a8a
    classDef merged fill:#dcfce7,stroke:#15803d,color:#14532d
    classDef closed fill:#e5e7eb,stroke:#6b7280,color:#1f2937
    classDef unverified fill:#f3f4f6,stroke:#9ca3af,color:#374151
    classDef best stroke-width:3px,stroke:#b45309
    classDef target stroke-width:3px,stroke:#4338ca
    I72580(["issue #72580 (open)"])
    subgraph Dup72601 ["PRs duplicating each other"]
        P72601["PR #72601 (open)"]
        P72614["PR #72614 (open)"]
        P73584["PR #73584 (open)"]
        P75207["PR #75207 (open)"]
    end
    P72601 -.->|partial| I72580
    class I72580 open
    class P72601 open
    class P72614 open
    class P73584 open
    class P75207 open
    class P72614 best
    class P75207 best
    class P72601 target
    click I72580 "https://github.com/NousResearch/hermes-agent/issues/72580"
    click P72601 "https://github.com/NousResearch/hermes-agent/pull/72601"
    click P72614 "https://github.com/NousResearch/hermes-agent/pull/72614"
    click P73584 "https://github.com/NousResearch/hermes-agent/pull/73584"
    click P75207 "https://github.com/NousResearch/hermes-agent/pull/75207"

Loading

Graph: solid arrow = fixes / best fix, dashed arrow = partial or unverified (see edge label); boxed group = PRs duplicating each other; amber border = best fix; indigo border = target; gray node = closed (state tag in the node label).

Cross-PR triage: Reviewed 4 pull requests and 1 issue in this complex. Each diff was read against this issue; Assessment working set: 34 kB of PR diffs, 7 kB of issue/PR text, 3 kB of discussion (5 comments), 7 verify verdicts. verdicts reflect diff content, not PR titles. Part of an automated triage batch.