feat(report): add hover explanations and polish run/hunt report UI

[jithin23-kv]

Problem

Both report renderers (opfor run and opfor hunt) had accumulated a set of UX and layout issues:

• Several metrics had no explanation for a first-time reader: severity badges look like they describe this result when they actually describe the evaluator's category, Confidence % is easy to mistake for a severity score, and hunt's Exploration stats (leads, depth, expand/dismiss) are internal jargon with zero context.
• The hunt report's "Reconnaissance" section mixed a genuinely live metric (probe count) with commander end-of-run synthesis (fingerprint, guardrails, weak points), so a run that ended before synthesis showed a confusing empty "recon found nothing" state.
• The Verifier row always displayed a model name (falling back to the commander's), even on runs that never used --verify — implying verification happened when it didn't.
• Several CSS layout bugs: exec-summary cells weren't bottom-aligned, single-turn transcripts had their text flush against the card edge (missing the turn-rail's implicit margin), the "View more/less details" toggle wasn't vertically centered, and the risk badge looked misaligned against the safety-score gauge (an optical illusion from font descenders, not an actual box-alignment bug).
• Stale/inconsistent copy: "Custom Suite" label, a leftover page footer duplicating info already in the cover band, and the hunt report's objective sitting in the cover header instead of the executive summary.

Solution

Added a lightweight, reusable hover-tooltip system to both report renderers — a small "i" icon that shows an explanation the instant the pointer enters the card/row (not the ~1s native title delay), rather than always-on text cluttering the layout. Used it to explain every metric that isn't self-evident, and fixed the CSS bugs found while building it (bottom alignment via space-between, an inline-flex vs flex bug that silently broke margin math on the details toggle, and the badge/gauge optical alignment). Also fixed the Verifier row to omit itself entirely rather than show a misleading fallback, and moved the misplaced guardrails/weak-points data into its own section instead of implying it's live recon.

Changes

Report renderers (core/src/report/render.ts, core/src/autonomous/report/html.ts):

• New hover-tooltip system: .info-icon + .info-tooltip/.info-hover, triggered by hovering the whole card/row, not just the icon.
• Explanatory tooltips added: Overall Verdict + risk badge (combined into one tooltip with a divider), Safety Score, Token Usage, Testing Cost, Duration, Findings (hunt), severity badges, Confidence, Attack Agents roles (hunt), Exploration stats (hunt).
• Fixed exec-summary cell bottom-alignment (space-between + body wrapper), single-turn transcript left margin (.no-rail), details-toggle vertical centering (display:flex, not inline-flex), and the verdict-badge/gauge optical misalignment.
• Cover band: run report gained a Judge Model column; hunt report now shows Commander/Operator/Scout models instead of Objective Outcome/Run Cost (both still shown elsewhere).
• Duration moved into the cover-date line, labeled "Ran for:".
• Standards now render as individual pills instead of one comma-joined string.
• Removed the redundant page footer (report ID/date already live in the cover band).
• Hunt report: objective moved into an "Objective" line at the bottom of the executive summary; Observed Guardrails/Weak Points moved out of "Reconnaissance" into a new "Guardrails & Weak Points" section after Findings, since that data is the commander's end-of-run synthesis, not something recon itself produces — the section is now omitted entirely (not shown empty) on runs that ended before synthesis.
• Verifier row is now omitted entirely when --verify wasn't used, instead of falling back to the commander's model.
• Removed the standalone attack-success-rate callouts in favor of Safety Score alone.

Suite-label rename (core/src/execute/runAll.ts, runAllBrowser.ts, core/src/execute/types.ts, runners/sdk/src/types.ts, runners/extension/popup.js): "Custom Suite" → "Custom" — three independent implementations of this label (Node run path, browser-extension path, and the extension's own display logic) all needed the same rename.

CLI flag rename (runners/cli/src/commands/hunt.ts, docs/hunt.md): --model → --commander-model for consistency with --operator-model/--scout-model.

Docs: core/src/autonomous/report/types.ts's verifierModel comment clarified to state it's only ever set when verification actually ran.

Issue

N/A

How to test

opfor run --config <path>            # then open the run report and hover the "i" icons
opfor hunt --endpoint <url> --objective "..." --verify   # then open the hunt report

Verified before every push: tsc --noEmit (core/cli/sdk), eslint, prettier --check, and npm test (277 passed / 0 failed) all clean, plus the full pre-commit pipeline (typecheck, lint, format, catalog build, skill validation, gitleaks) passed on commit.

Screenshots

Verified visually via headless-Chrome screenshots at each step (exec-strip cell alignment, single-turn transcript margin, combined verdict tooltip with divider, severity/confidence tooltips rendering inline) rather than by inspection alone — not attached here, but every layout claim above was screenshot-checked before merging into this branch.

Summary by CodeRabbit

• New Features

  • Enhanced HTML reports with clearer Pass/Fail verdicts, risk explanations, evaluator tooltips, safety scores, findings, costs, durations, and model details.
  • Added objective callouts and dedicated sections for recon guardrails and weak points.
  • Improved responsive transcript layouts, standards presentation, and overall report readability.

• Updates

  • Custom evaluation runs are now labeled “Custom.”
  • Renamed the hunt CLI option from --model to --commander-model.

• Documentation

  • Updated hunt command guidance and model configuration documentation to reflect the new option name.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 43145f84-8a30-4f8b-8f3f-171297827e57

📥 Commits

Reviewing files that changed from the base of the PR and between e2b558d and e2a1947.

📒 Files selected for processing (2)

• core/src/autonomous/report/html.ts
• core/src/report/render.ts

🚧 Files skipped from review as they are similar to previous changes (2)

• core/src/report/render.ts
• core/src/autonomous/report/html.ts

---

Walkthrough

The pull request redesigns autonomous and shared HTML reports, updates report metadata and transcript behavior, standardizes non-suite labels as Custom, and renames the hunt commander model option to --commander-model.

Changes

Report presentation

Layer / File(s)	Summary
Autonomous report redesign core/src/autonomous/report/html.ts, core/src/autonomous/report/types.ts	The autonomous report uses Fail/Pass verdicts, risk explanations, model metadata, revised executive cards, tooltips, and a dedicated Guardrails & Weak Points section.
Shared report renderer updates core/src/report/render.ts	The shared renderer adds risk and severity explanations, standards pills, revised summary cards, duration metadata, and updated transcript behavior.
Hunt option and suite label consistency runners/cli/src/commands/hunt.ts, docs/hunt.md, core/src/execute/*, runners/extension/popup.js, runners/sdk/src/types.ts	The commander option is renamed to --commander-model. Non-suite evaluator selections use the Custom label across reports and documentation.

Estimated code review effort: 3 (Moderate) | ~25 minutes

Possibly related PRs

• KeyValueSoftwareSystems/agent-opfor#227: Updates related suite-labeling and report presentation behavior.
• KeyValueSoftwareSystems/agent-opfor#230: Modifies the shared report renderer, transcript behavior, risk tooltips, and standards presentation.
• KeyValueSoftwareSystems/agent-opfor#233: Introduces related autonomous report and metadata changes.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the main report UI improvements and hover explanations.
Description check	✅ Passed	The description includes all required sections and provides detailed problem, solution, changes, testing, and screenshot information.
Docstring Coverage	✅ Passed	Docstring coverage is 80.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[coderabbitai]

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 `@core/src/autonomous/report/html.ts`:
- Around line 762-763: Update the Safety Score tooltip in
core/src/autonomous/report/html.ts lines 762-763 to say “scoreable findings”
instead of “scoreable threads,” and update the Findings tooltip at lines 771-772
to say “One entry per finding” instead of “One entry per attack thread.”

🪄 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: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 6beba5fa-7f74-4864-8ed2-bcbab4c063c2

📥 Commits

Reviewing files that changed from the base of the PR and between a67cca9 and 7953623.

📒 Files selected for processing (10)

• core/src/autonomous/report/html.ts
• core/src/autonomous/report/types.ts
• core/src/execute/runAll.ts
• core/src/execute/runAllBrowser.ts
• core/src/execute/types.ts
• core/src/report/render.ts
• docs/hunt.md
• runners/cli/src/commands/hunt.ts
• runners/extension/popup.js
• runners/sdk/src/types.ts