adversarial-review-gate 2.0.3 → 2.1.0

package/.claude-plugin/marketplace.json

@@ -1,7 +1,7 @@
 {
   "name": "adversarial-review",
   "owner": {
-    "name": "Khoa Pham"
+    "name": "louisphamdev"
   },
   "metadata": {
     "description": "Internal marketplace hosting the adversarial-review code-quality gate."
@@ -10,7 +10,7 @@
     {
       "name": "adversarial-review",
       "source": "./",
-      "description": "Soft Stop-hook gate that forces an adversarial review of significant code changes before an agent finishes."
+      "description": "Stop-hook gate (default: enforced) that forces an adversarial review of significant code changes before an agent finishes."
     }
   ]
 }

package/.claude-plugin/plugin.json

@@ -1,9 +1,9 @@
 {
   "name": "adversarial-review",
-  "version": "2.0.0",
+  "version": "2.1.0",
   "description": "NodeJS multi-tool adversarial review gate. Installs a SessionStart baseline hook and a Stop gate hook for Claude Code. Supports multiple host tools (Claude Code, Codex, opencode) with configurable reviewer mappings, policy modes (soft/enforced/strict-ci), and native or wrapper enforcement. Significant code changes must pass an adversarial review before the agent finishes.",
   "author": {
-    "name": "adversarial-review contributors"
+    "name": "louisphamdev"
   },
   "keywords": ["code-review", "quality", "hook", "stop-hook", "adversarial"],
   "hooks": {

package/CHANGELOG.md

@@ -0,0 +1,78 @@
+# Changelog
+
+All notable changes to this project are documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [2.1.0] - 2026-06-13
+
+Public-release perfection pass: hardening, machine-wide install, and docs.
+
+### Added
+- `install --global` (alias `--user`): machine-wide install that writes the
+  host/reviewer defaults to `~/.adversarial-review/config.json` and merges the
+  Claude Code `SessionStart` + `Stop` hooks into the user-level
+  `~/.claude/settings.json`.
+- `uninstall` command (`uninstall [--user]`): removes the hooks this tool wrote
+  and the install-registry entry, for both project and machine-wide installs.
+- `install` now merges into an existing Claude Code `settings.json` (preserving
+  other keys) instead of replacing it.
+- Additional `doctor` checks for the merged settings, the opencode read-only
+  agent, and reviewer isolation.
+- CI workflow (`.github/workflows/ci.yml`): tests on
+  `ubuntu-latest`/`windows-latest` x Node 20/22, plus `npm run pack:dry-run`.
+- `SECURITY.md`, `CONTRIBUTING.md`, and this `CHANGELOG.md`.
+- README badges (npm version, CI, license, node) and documentation for the new
+  install flags and `uninstall`.
+
+### Changed
+- The installed Claude Code `Stop` hook now carries a **300-second timeout** so a
+  debate-tier review is not aborted mid-run.
+- Hardened reviewer isolation: enforced/strict-ci modes reject any reviewer that
+  is not `readOnly && noEdit`.
+- More robust coverage parsing in verdict handling.
+- Deduplication of changed-file scope so a file is not reviewed twice.
+- `package.json`: added `prepublishOnly` (test + pack dry-run) and
+  `publishConfig.access = public`; added `CHANGELOG.md` to the files allowlist.
+
+### Fixed
+- `skills/adversarial-review-setup/SKILL.md` referenced the wrong package name
+  (`adversarial-review` instead of `adversarial-review-gate`).
+
+## [2.0.3] - 2026-06-13
+
+### Fixed
+- A fresh install now produces a working `enforced` + opencode gate out of the
+  box: the installer creates the read-only opencode `adversarial-reviewer` agent
+  (idempotent), writes `reviewers.opencode.readOnlyConfig: true`, and skips the
+  install-time agent-existence check so a clean machine can bootstrap.
+
+## [2.0.2] - 2026-06-13
+
+### Added
+- A package-name-matching `adversarial-review-gate` bin so `npx
+  adversarial-review-gate` resolves correctly.
+
+## [2.0.1] - 2026-06-13
+
+### Fixed
+- Installer bin name corrected to `adversarial-review-gate`.
+
+## [2.0.0] - 2026-06-13
+
+### Added
+- Initial NodeJS multi-tool adversarial-review gate, replacing the previous
+  Python/Claude-plugin implementation.
+- Multi-host support (Claude Code native Stop hook; codex, opencode,
+  github-copilot-cli, antigravity via wrapper) with configurable reviewer
+  mappings and self-review (`none`) orchestration.
+- Policy modes `soft` / `enforced` / `strict-ci`, layered config
+  (default < user < project) with a tighten-only user policy floor.
+- `install`, `check`, `run`, `doctor`, and `hook` commands.
+
+[2.1.0]: https://github.com/louisphamdev/adversarial-review/releases/tag/v2.1.0
+[2.0.3]: https://github.com/louisphamdev/adversarial-review/releases/tag/v2.0.3
+[2.0.2]: https://github.com/louisphamdev/adversarial-review/releases/tag/v2.0.2
+[2.0.1]: https://github.com/louisphamdev/adversarial-review/releases/tag/v2.0.1
+[2.0.0]: https://github.com/louisphamdev/adversarial-review/releases/tag/v2.0.0

package/README.md

@@ -1,5 +1,10 @@
 # adversarial-review
 
+[![npm version](https://img.shields.io/npm/v/adversarial-review-gate.svg)](https://www.npmjs.com/package/adversarial-review-gate)
+[![CI](https://github.com/louisphamdev/adversarial-review/actions/workflows/ci.yml/badge.svg)](https://github.com/louisphamdev/adversarial-review/actions/workflows/ci.yml)
+[![license](https://img.shields.io/npm/l/adversarial-review-gate.svg)](./LICENSE)
+[![node](https://img.shields.io/node/v/adversarial-review-gate.svg)](https://nodejs.org)
+
 A NodeJS multi-tool adversarial review gate for coding agents.
 
 The gate stops a coding agent from finishing a turn when a significant code
@@ -29,20 +34,62 @@ npx adversarial-review-gate install \
 
 The installer detects available host and reviewer tools, verifies each reviewer
 binary (it must resolve on `PATH` and pass a version/auth check), and writes the
-project config plus any native host integration files. Run with no flags for the
-interactive wizard, or pass `--hosts`/`--reviewer` for a scripted setup.
+project config plus any native host integration files.
+
+`--hosts` and `--reviewer` are **required** — there is no interactive wizard
+(one may be added in a future release). Pass them explicitly for a scripted
+setup.
+
+What the installer writes for you (idempotent — it never clobbers a file you
+have customized):
+
+- The **project config** at `.adversarial-review/config.json` with the
+  host → reviewer mapping you chose.
+- For an **opencode** reviewer: the read-only `adversarial-reviewer` opencode
+  agent at `~/.config/opencode/agent/adversarial-reviewer.md` (skipped if it
+  already exists) **and** `reviewers.opencode.readOnlyConfig: true` in the
+  project config, so enforced-mode isolation passes out of the box.
+- For **Claude Code**: the native `SessionStart` + `Stop` hooks in
+  `.claude/settings.json`.
+- The user-level install registry at `~/.adversarial-review/install.json`.
 
 Supported flags (see `src/cli/install.js`):
 
 | Flag | Meaning |
 |---|---|
-| `--hosts a,b` | Comma-separated list of hosts to install (repeatable). |
-| `--reviewer host=reviewer` | Reviewer mapping for a host (repeatable). Use `host=none` for self-review. |
+| `--hosts a,b` | Comma-separated list of hosts to install (repeatable). **Required.** |
+| `--reviewer host=reviewer` | Reviewer mapping for a host (repeatable). Use `host=none` for self-review. **Required per host.** |
+| `--global` / `--user` | Machine-wide install: write the defaults to `~/.adversarial-review/config.json` and merge the Claude Code hooks into your user-level `~/.claude/settings.json` instead of the per-project files. |
 | `--dry-run` | Print every planned write and exit 0 without writing anything. |
 | `--project-config <path>` | Write the project config to an explicit path. |
 
-> There is no `--user-config` flag. The machine-wide defaults file below is
-> written/edited by hand — the installer does not generate it.
+### Machine-wide install
+
+To install once for **every** project on the machine, add `--global` (alias
+`--user`):
+
+```bash
+npx adversarial-review-gate install --global \
+  --hosts claude-code,codex \
+  --reviewer claude-code=opencode \
+  --reviewer codex=opencode
+```
+
+This writes the host/reviewer defaults to `~/.adversarial-review/config.json`
+and merges the Claude Code `SessionStart` + `Stop` hooks into your user-level
+`~/.claude/settings.json` (existing keys are preserved). New projects then
+inherit the gate without re-running install per project.
+
+### Uninstall
+
+```bash
+npx adversarial-review-gate uninstall          # remove the project install
+npx adversarial-review-gate uninstall --user   # remove the machine-wide install
+```
+
+`uninstall` removes the hooks this tool wrote (from the project or user
+`settings.json`) and the install-registry entry. Re-run `doctor` afterward to
+confirm the gate is no longer active.
 
 ### After install
 
@@ -55,13 +102,25 @@ project config validity, and the Claude Code session baseline.
 
 ### Machine-wide defaults
 
-A user-level `~/.adversarial-review/config.json` provides host/reviewer defaults
-that apply across **all** projects. Config is layered in this order, where each
-later layer overrides the earlier ones — except the policy floor, which can only
-ever tighten, never loosen:
+Two distinct user-level files shape every project on the machine:
+
+- **`~/.adversarial-review/config.json`** — the user **override layer**. It
+  provides host/reviewer defaults and policy that apply across **all** projects.
+  As a normal config layer it can either loosen **or** tighten relative to the
+  built-in defaults; a project config can in turn override it.
+- **`~/.adversarial-review/policy.json`** — the **policy floor**. It is
+  **tighten-only**: it can raise the minimum policy (e.g. force `enforced` or
+  `strict-ci`) but no later layer — user config or project config — can ever
+  loosen below it.
+
+Config is layered in this order, where each later layer overrides the earlier
+ones, and the policy floor is applied last and can only ever tighten:
 
 ```text
-DEFAULT_CONFIG  <  userConfig (~/.adversarial-review/config.json)  <  projectConfig (.adversarial-review/config.json)  <  policy floor
+DEFAULT_CONFIG
+  <  userConfig    (~/.adversarial-review/config.json)   # override layer (loosen or tighten)
+  <  projectConfig (.adversarial-review/config.json)     # override layer (loosen or tighten)
+  <  policyFloor   (~/.adversarial-review/policy.json)   # tighten-only, applied last
 ```
 
 Example `~/.adversarial-review/config.json`:
@@ -87,7 +146,8 @@ Example `~/.adversarial-review/config.json`:
 
 With this in place, a new project inherits the host/reviewer mapping and the
 `enforced` mode without re-running install per project. A project may still ship
-its own `.adversarial-review/config.json` to make policy **stricter** (see
+its own `.adversarial-review/config.json` to override these defaults, but it can
+never go below the policy floor in `~/.adversarial-review/policy.json` (see
 [Policy Modes](#policy-modes)).
 
 ---
@@ -135,9 +195,23 @@ explicitly `none`.
 Claude Code        -> codex        (external reviewer)
 Codex              -> opencode     (external reviewer)
 opencode           -> none         (self-review orchestration)
-GitHub Copilot CLI -> claude-code  (external reviewer, if available)
 ```
 
+The five hosts in the registry, with their enforcement and whether they can
+delegate to an **external** reviewer:
+
+| Host | Enforcement | External reviewer? |
+|---|---|---|
+| `claude-code` | native-enforced | yes |
+| `codex` | wrapper-enforced | yes |
+| `opencode` | wrapper-enforced | yes |
+| `github-copilot-cli` | wrapper-enforced | no — self-review (`none`) only |
+| `antigravity` | wrapper-enforced | no — self-review (`none`) only |
+
+`github-copilot-cli` and `antigravity` are marked `supportsExternalReview: false`
+in the registry, so they cannot be mapped to an external reviewer — use
+`--reviewer github-copilot-cli=none` (self-review orchestration) for those.
+
 Reviewer tools are verified during install: binary must exist, basic version
 check must succeed, and auth check must pass where available.
 
@@ -162,38 +236,50 @@ unresolved Critical or Important findings.
 
 ## Using opencode as the reviewer (read-only)
 
-**This setup is required before opencode can pass the gate.** It was a real
-gotcha during local validation, so follow every point below.
+**The installer sets this up for you.** When you map any host to
+`--reviewer <host>=opencode`, `install` writes a working read-only opencode
+reviewer with no manual steps:
+
+- It creates the bundled `adversarial-reviewer` agent at
+  `~/.config/opencode/agent/adversarial-reviewer.md` (idempotent — it is
+  **skipped if the file already exists**, so a customized agent is never
+  overwritten).
+- It writes `reviewers.opencode.readOnlyConfig: true` into the project config so
+  the gate's enforced-mode isolation check passes.
 
 opencode is invoked as `opencode run --pure --agent adversarial-reviewer -f <diff>`
-with the review brief delivered on stdin. For that to work, opencode must have an
-`adversarial-reviewer` agent defined, for example at
-`~/.config/opencode/agent/adversarial-reviewer.md`.
+with the review brief delivered on stdin.
+
+### What the bundled setup guarantees
 
-Three hard requirements:
+The agent the installer ships satisfies three invariants that the gate enforces.
+You do not configure these by hand — verify them with
+`npx adversarial-review-gate doctor` and `opencode agent list`:
 
-1. **The agent MUST be `mode: primary` — NOT `subagent`.** `opencode run --agent`
+1. **The agent is `mode: primary` — NOT `subagent`.** `opencode run --agent`
    rejects a subagent and **silently** falls back to the full-permission default
    agent, printing `Falling back to default agent` to stderr. The gate detects
    that marker and rejects the review as an operational failure
    (`reviewer_agent_fallback`), so a subagent-mode agent can never pass — even if
    it printed a perfect verdict block.
 
-2. **It must be read-only.** Set `permission` to deny everything and turn tools
-   off, so the gate's enforced isolation check passes. In `enforced`/`strict-ci`
-   the gate refuses any reviewer whose `verify()` capabilities are not
-   `readOnly === true && noEdit === true` (`reviewer_not_isolated`). The opencode
-   adapter only asserts those capabilities when
-   `reviewers.opencode.readOnlyConfig: true` is set in config — so you must both
-   make the agent read-only **and** set that flag.
+2. **The agent is read-only.** `permission` denies everything and tools are
+   turned off, and `reviewers.opencode.readOnlyConfig: true` is set, so the
+   adapter reports `readOnly === true && noEdit === true`. In
+   `enforced`/`strict-ci` the gate refuses any reviewer whose `verify()`
+   capabilities are not isolated (`reviewer_not_isolated`).
 
-3. **The agent body must contain the verdict-block format the gate parses.** The
+3. **The agent body contains the verdict-block format the gate parses.** The
    brief on stdin carries the per-job `job_id` / `diff_hash` / `payload_hash` /
-   `reviewer` / `level`; the agent must echo those exact values back inside a
-   single `<<<ADVERSARIAL-REVIEW-VERDICT>>> ... <<<END>>>` block (see
+   `reviewer` / `level`; the agent echoes those exact values back inside a single
+   `<<<ADVERSARIAL-REVIEW-VERDICT>>> ... <<<END>>>` block (see
    [Verdict Format](#verdict-format)) with nothing after `<<<END>>>`.
 
-Minimal `~/.config/opencode/agent/adversarial-reviewer.md`:
+### If you customize the agent, keep these invariants
+
+Because the installer never overwrites an existing agent file, an edited
+`~/.config/opencode/agent/adversarial-reviewer.md` must still satisfy all three
+invariants above. A minimal shape:
 
 ```markdown
 ---
@@ -274,11 +360,20 @@ node C:\abs\path\to\adversarial-review\bin\adversarial-review.js run --host code
 
 ## Claude Code (native)
 
-Claude Code is the only **native-enforced** host. The installer adds two hooks to
-`.claude/settings.json`:
+Claude Code is the only **native-enforced** host. **The installer writes the
+hooks for you** — when `claude-code` is in `--hosts`, `install` merges two hooks
+into `.claude/settings.json` (or, with `--global`, into your user-level
+`~/.claude/settings.json`):
 
 - A **SessionStart** hook that records the workspace baseline.
-- A **Stop** hook that applies the gate before the turn finishes.
+- A **Stop** hook (with a **300-second timeout**) that applies the gate before
+  the turn finishes.
+
+### What the bundled setup guarantees
+
+The hook commands invoke `adversarial-review-gate` directly when it resolves on
+`PATH` (a global npm install), otherwise via `npx adversarial-review-gate`. The
+written block looks like this:
 
 ```json
 {
@@ -288,7 +383,7 @@ Claude Code is the only **native-enforced** host. The installer adds two hooks t
         "hooks": [
           {
             "type": "command",
-            "command": "node /abs/path/to/adversarial-review/bin/adversarial-review.js hook --host claude-code --event session-start"
+            "command": "adversarial-review-gate hook --host claude-code --event session-start"
           }
         ]
       }
@@ -298,7 +393,8 @@ Claude Code is the only **native-enforced** host. The installer adds two hooks t
         "hooks": [
           {
             "type": "command",
-            "command": "node /abs/path/to/adversarial-review/bin/adversarial-review.js hook --host claude-code --event stop"
+            "command": "adversarial-review-gate hook --host claude-code --event stop",
+            "timeout": 300
           }
         ]
       }
@@ -307,15 +403,19 @@ Claude Code is the only **native-enforced** host. The installer adds two hooks t
 }
 ```
 
-For a **local (non-marketplace) install**, the hook command needs an **absolute
-path** to `bin/adversarial-review.js`. `${CLAUDE_PLUGIN_ROOT}` only resolves
-inside a marketplace plugin, so it will not work for a plain local checkout — use
-the absolute node path as shown above.
+> The Stop hook carries a **300s timeout** because an external review can take up
+> to a few minutes — Claude Code will not abort the hook before the review
+> finishes.
 
 > Both hooks are required. A Stop hook that sees edit evidence but finds **no
 > recorded SessionStart baseline** fails closed (blocks) in `enforced`/`strict-ci`,
-> because the full change scope is unknown. Restart Claude Code after editing
-> `settings.json`.
+> because the full change scope is unknown. **Restart Claude Code after install**
+> so it re-reads `settings.json`. Verify the hooks with
+> `npx adversarial-review-gate doctor`.
+
+> For a **local (non-marketplace) checkout** where the package is not on `PATH`,
+> the hook command needs an **absolute path** to `bin/adversarial-review.js`;
+> `${CLAUDE_PLUGIN_ROOT}` only resolves inside a marketplace plugin.
 
 ---
 
@@ -337,7 +437,9 @@ never committed.
 
 ## Cost
 
-An external opencode review takes roughly **30 seconds** and **BLOCKS in
+An external opencode review takes roughly **30 seconds** — and a debate-tier
+review on a large or sensitive diff can take **up to a few minutes** (the Claude
+Code Stop hook is given a 300-second timeout for this reason). It **BLOCKS in
 `enforced`** until it passes. With a machine-wide `enforced` config, that gate
 runs on every significant-edit Stop across **all** projects — which adds up
 quickly.
@@ -563,14 +665,19 @@ Common issues:
 ## Commands
 
 ```bash
-npx adversarial-review-gate install    # Interactive install wizard
 npx adversarial-review-gate install --hosts claude-code,codex --reviewer claude-code=opencode --reviewer codex=opencode
-npx adversarial-review-gate install --dry-run  # Preview without writing
+npx adversarial-review-gate install --global --hosts claude-code --reviewer claude-code=opencode  # Machine-wide
+npx adversarial-review-gate install --dry-run ...  # Preview without writing
+npx adversarial-review-gate uninstall          # Remove the project install
+npx adversarial-review-gate uninstall --user   # Remove the machine-wide install
 npx adversarial-review-gate check      # Run the gate manually against current working tree
 npx adversarial-review-gate run --host codex -- codex exec "..."  # Wrapper mode
 npx adversarial-review-gate doctor     # Verify installation
 ```
 
+`--hosts` and `--reviewer` are required for `install`; there is no interactive
+wizard (one may be added later).
+
 For a local (unpublished) checkout, `npx adversarial-review-gate` becomes
 `node /abs/path/to/adversarial-review/bin/adversarial-review.js`.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "adversarial-review-gate",
-  "version": "2.0.3",
+  "version": "2.1.0",
   "description": "NodeJS multi-tool adversarial review gate for coding agents.",
   "type": "module",
   "bin": {
@@ -10,13 +10,18 @@
   "scripts": {
     "test": "node --test",
     "doctor": "node ./bin/adversarial-review.js doctor --dry-run",
-    "pack:dry-run": "npm pack --dry-run"
+    "pack:dry-run": "npm pack --dry-run",
+    "prepublishOnly": "npm test && npm run pack:dry-run"
+  },
+  "publishConfig": {
+    "access": "public"
   },
   "files": [
     "bin/",
     "src/",
     ".claude-plugin/",
     "README.md",
+    "CHANGELOG.md",
     "LICENSE"
   ],
   "engines": {

package/src/cli/doctor.js

@@ -11,31 +11,44 @@
 import { readFile } from "node:fs/promises";
 import { existsSync } from "node:fs";
 import path from "node:path";
-import os from "node:os";
 import { fileURLToPath } from "node:url";
 
 import { HOSTS } from "../hosts/index.js";
 import { createReviewer } from "../reviewers/index.js";
-import { loadEffectiveConfig } from "../core/load-config.js";
+import { loadEffectiveConfig, resolveHomeDir } from "../core/load-config.js";
+import { detectClaudeCodeHooks, claudeCodeSettingsPath } from "../hosts/claude-code.js";
 
 // Paths relative to home / cwd.
 const PROJECT_CONFIG_REL = path.join(".adversarial-review", "config.json");
 const USER_CONFIG_REL = path.join(".adversarial-review", "config.json");
 const USER_POLICY_REL = path.join(".adversarial-review", "policy.json");
+const OPENCODE_AGENT_REL = path.join(
+  ".config",
+  "opencode",
+  "agent",
+  "adversarial-reviewer.md"
+);
 
 // ---------------------------------------------------------------------------
 // Helpers
 // ---------------------------------------------------------------------------
 
-/** Resolve home from env, falling back to os.homedir(). Honors
- * ADVERSARIAL_REVIEW_HOME so doctor reports the SAME user-level base that
- * loadEffectiveConfig (the gate's loader) uses. */
-function homeDir(env) {
-  if (env) {
-    const fromEnv = env.ADVERSARIAL_REVIEW_HOME || env.HOME || env.USERPROFILE;
-    if (fromEnv) return fromEnv;
+/**
+ * Read + tolerantly parse a Claude Code settings.json. Returns {} on any error
+ * (missing or corrupt) so the hook-presence detection never throws.
+ *
+ * @param {string} filePath
+ * @returns {Promise<object>}
+ */
+async function readSettingsTolerant(filePath) {
+  try {
+    const raw = await readFile(filePath, "utf8");
+    const parsed = JSON.parse(raw);
+    if (parsed && typeof parsed === "object" && !Array.isArray(parsed)) return parsed;
+    return {};
+  } catch {
+    return {};
   }
-  return os.homedir();
 }
 
 /** Read package.json to get the package version. */
@@ -77,7 +90,7 @@ export async function doctorCommand(argv, io) {
   const json = argv.includes("--json");
   const cwd = io.cwd || process.cwd();
   const env = io.env || process.env;
-  const home = homeDir(env);
+  const home = resolveHomeDir(env);
 
   // Read package version.
   const version = await readPackageVersion();
@@ -110,6 +123,23 @@ export async function doctorCommand(argv, io) {
   // hosts.)
   const effectiveConfig = await loadEffectiveConfig(cwd, io);
 
+  // For native hosts (claude-code) detect whether OUR SessionStart + Stop hooks
+  // are actually registered in the relevant .claude/settings.json — at BOTH
+  // project (<cwd>/.claude) and user (<home>/.claude) scope. A configured host
+  // whose hooks are missing means the gate would never fire.
+  const projectSettingsPath = claudeCodeSettingsPath(cwd);
+  const userSettingsPath = claudeCodeSettingsPath(home);
+  const projectHooks = detectClaudeCodeHooks(
+    await readSettingsTolerant(projectSettingsPath)
+  );
+  const userHooks = detectClaudeCodeHooks(
+    await readSettingsTolerant(userSettingsPath)
+  );
+
+  // opencode read-only agent presence (machine-wide, under <home>).
+  const opencodeAgentPath = path.join(home, OPENCODE_AGENT_REL);
+  const opencodeAgentExists = existsSync(opencodeAgentPath);
+
   // Enumerate configured hosts.
   const configuredHostIds = Object.keys(effectiveConfig.hosts || {});
   const hostReports = [];
@@ -135,6 +165,54 @@ export async function doctorCommand(argv, io) {
       reviewerNote: reviewerResult.note || null,
       reviewerReason: reviewerResult.reason || null,
     };
+
+    // For claude-code, report whether our native hooks are registered. We treat
+    // the host as "registered" if BOTH SessionStart + Stop are present at EITHER
+    // scope (project or user), since a user-scope install also covers this cwd.
+    if (hostId === "claude-code") {
+      const projectRegistered = projectHooks.sessionStart && projectHooks.stop;
+      const userRegistered = userHooks.sessionStart && userHooks.stop;
+      hostReport.hooks = {
+        projectSettingsPath,
+        userSettingsPath,
+        project: projectHooks,
+        user: userHooks,
+        registered: projectRegistered || userRegistered,
+      };
+      if (!hostReport.hooks.registered) {
+        warnings.push(
+          `WARNING: Host "claude-code" is configured but our SessionStart + Stop ` +
+            `hooks are NOT registered in ${projectSettingsPath} (project) or ` +
+            `${userSettingsPath} (user). Run \`adversarial-review install\` to register them.`
+        );
+      }
+    }
+
+    // For an opencode reviewer, report whether the read-only agent exists and
+    // whether reviewers.opencode.readOnlyConfig is set — the two settings the
+    // README promises and that enforced-mode isolation depends on.
+    if (reviewerId === "opencode") {
+      const readOnlyConfig =
+        effectiveConfig.reviewers?.opencode?.readOnlyConfig === true;
+      hostReport.opencode = {
+        agentPath: opencodeAgentPath,
+        agentExists: opencodeAgentExists,
+        readOnlyConfig,
+      };
+      if (!opencodeAgentExists) {
+        warnings.push(
+          `WARNING: opencode reviewer for host "${hostId}" but the read-only agent ` +
+            `is MISSING at ${opencodeAgentPath}. Run \`adversarial-review install\` to create it.`
+        );
+      }
+      if (!readOnlyConfig) {
+        warnings.push(
+          `WARNING: reviewers.opencode.readOnlyConfig is not set for host "${hostId}". ` +
+            `Enforced-mode isolation (readOnly && noEdit) will fail without it.`
+        );
+      }
+    }
+
     hostReports.push(hostReport);
 
     // Warn about wrapper/advisory hosts.
@@ -247,6 +325,25 @@ function printHumanReport(report, io) {
       } else {
         w(`    reviewer status: UNAVAILABLE (${h.reviewerReason || "unknown"})\n`);
       }
+      // Native hook registration (claude-code only).
+      if (h.hooks) {
+        w(
+          `    native hooks registered: ${h.hooks.registered ? "yes" : "NO"}\n`
+        );
+        w(
+          `      project (${h.hooks.projectSettingsPath}): ` +
+            `SessionStart=${h.hooks.project.sessionStart} Stop=${h.hooks.project.stop}\n`
+        );
+        w(
+          `      user (${h.hooks.userSettingsPath}): ` +
+            `SessionStart=${h.hooks.user.sessionStart} Stop=${h.hooks.user.stop}\n`
+        );
+      }
+      // opencode reviewer details.
+      if (h.opencode) {
+        w(`    opencode agent exists: ${h.opencode.agentExists ? "yes" : "NO"} (${h.opencode.agentPath})\n`);
+        w(`    opencode readOnlyConfig: ${h.opencode.readOnlyConfig}\n`);
+      }
     }
   }