@polderlabs/bizar 9.0.0 → 10.0.5

package/.claude/worktrees/kanban-overhaul/.claude/commands/team.md

@@ -1,132 +0,0 @@
----
-description: Spawn a Claude Code agent team — coordinate Thor (M2.7), Tyr (M3), Mimir (research), and Hermod (git) to complete a complex mission end-to-end.
-allowed-tools: Read, Bash, Agent
----
-
-# Team — Spawn a Claude Code Agent Team
-
-The `/team` command spawns a **coordinated agent team** for a complex,
-multi-step mission. The lead agent coordinates teammates; each teammate
-owns a disjoint scope; results are synthesized and committed.
-
-The full mission is available as `$ARGUMENTS` (or composed from `$1`).
-
-## When to Use `/team`
-
-Use `/team` for any non-trivial mission that benefits from **parallel
-work + explicit coordination**. Examples:
-
-- **Refactor** — split a module across Thor (M2.7) and Tyr (M3) in parallel.
-- **Multi-feature build** — new feature A (Thor) + new feature B (Tyr) + tests (Heimdall).
-- **Research + implementation** — Mimir researches, Thor implements, Forseti audits.
-- **Migration / upgrade** — Hermod handles the git/PR lifecycle while Thor/Tyr write code.
-- **Bug hunt** — Mimir traces the data flow while Tyr writes the fix and tests.
-
-Do NOT use `/team` for trivial single-file edits or single-question
-clarifications. Use `@heimdall` directly for those.
-
-## Default Team Composition
-
-If the user does not specify a team composition, spawn this default:
-
-| Role          | Agent   | Model             | Job                                   |
-|---------------|---------|-------------------|---------------------------------------|
-| Lead          | @odin   | minimax/MiniMax-M3   | Coordinate, synthesize, gate quality  |
-| Implementer 1 | @thor   | minimax/MiniMax-M2.7 | Moderate-complexity implementation   |
-| Implementer 2 | @tyr    | minimax/MiniMax-M3   | Complex / cross-cutting work         |
-| Researcher    | @mimir  | 9router/kr/auto  | Codebase research, pattern discovery  |
-| Git ops       | @hermod | minimax/MiniMax-M2.7 | Branch, commit, push, PR             |
-| Reviewer      | @forseti| minimax/MiniMax-M3   | Audit plan + final output            |
-
-You can swap or remove any of these — but always keep at least one
-implementer and a reviewer.
-
-## Protocol (How Odin Runs a `/team` Invocation)
-
-1. **Read `.bizar/PROJECT.md`** if it exists. Factor the project
-   description into the team mission.
-2. **Read `bizar memory status`**. If the vault is reachable, run
-   `bizar memory search "<topic>"` for prior context on the mission.
-3. **Decompose** the mission into disjoint work items. Each item
-   names:
-   - The agent that owns it
-   - The file scope (paths/globs)
-   - The deliverable (a function, a test, a commit, a doc)
-   - The exit signal ("done when X is committed")
-4. **Launch in parallel** — use the **Agent tool** in a **single
-   message** with multiple `Agent` calls so all teammates run
-   concurrently. Each call's prompt names the Bizar agent
-   (`thor`, `tyr`, `mimir`, `hermod`, `forseti`) so the audit
-   trail stays readable. For long-running work, pass
-   `run_in_background: true`.
-5. **Monitor** — while the team runs, return control to the user.
-   The team runs asynchronously. Check teammate progress only
-   when the user asks (use `SendMessage` to the named agent).
-6. **Synthesize** — when teammates report back, gate through @forseti
-   (audit) and @hermod (commit/push). Then report the final outcome
-   to the user with file:line references.
-
-## Pre-Dispatch Checklist (MANDATORY)
-
-- [ ] Each teammate has a **disjoint file scope** — no two agents
-      edit the same file or directory
-- [ ] Lockfiles, `package.json`, root configs, `tsconfig.json`,
-      `vite.config.*`, `Dockerfile`, CI configs are assigned to ONE
-      agent or marked READ-ONLY for everyone else
-- [ ] No subagent has both `bash: allow` AND a write-level git task
-      in the same batch — Hermod is the only git writer
-- [ ] Each teammate's scope is named in plain English
-- [ ] You have prepended the **Sibling-Awareness Block** to every
-      prompt — see Odin baseline, "Parallel Dispatch Coordination"
-- [ ] You have called `bizar memory status` and confirmed the vault
-      is reachable
-
-## Example Mission Strings
-
-**Refactor a module:**
-```
-Refactor src/api/auth.ts into a clean module: extract JWT validation
-into src/auth/jwt.ts, rate-limiting into src/auth/rate-limit.ts, and
-keep the public surface stable. Add tests for each new module. Update
-docs/architecture.md with the new module map. Do not change the HTTP
-route handlers.
-```
-
-**Multi-feature build:**
-```
-Add a kanban board: (1) Thor — backend API at /api/tasks with CRUD
-endpoints and SQLite persistence. (2) Tyr — frontend React component
-in src/components/Kanban.tsx with 5 columns. (3) Heimdall — write
-the OpenAPI spec at docs/api/tasks.yaml. (4) Hermod — open a PR
-named feat/kanban-board.
-```
-
-**Bug hunt:**
-```
-Mimir — trace the data flow from POST /api/checkout through to the
-Stripe webhook handler; find any place where the order status is
-written before the payment is confirmed. Tyr — write a regression
-test that reproduces the race. Thor — fix the bug. Forseti —
-audit the fix for completeness. Hermod — commit + push.
-```
-
-## Tool Reference (Claude Code)
-
-- **Agent tool** — spawn one subagent per call. Use multiple
-  `Agent` calls in a single message for parallelism.
-- `run_in_background: true` on the Agent tool — for tasks that
-  span more than a few minutes.
-- `SendMessage` — talk to a running background agent by name.
-
-## Failure Modes
-
-- **Two agents write the same file** — STOP, revert one branch, reassign.
-- **One agent stalls** — use `TaskStop` to terminate, route the
-  remaining scope to another agent.
-- **Tests fail after team work** — dispatch `thor` to investigate;
-  do not paper over the failure.
-- **Git conflict** — route to `hermod` for `git rebase` / merge
-  resolution.
-- **A teammate goes silent** — send a `SendMessage` to the named
-  agent for a status report; if dead, kill and re-spawn with a
-  fresh prompt.

package/.claude/worktrees/kanban-overhaul/.claude/commands/test.md

@@ -1,62 +0,0 @@
----
-description: Run the project's test suite via the Bizar test gate. Auto-detects jest/vitest/bun/pytest/cargo/go. Exits non-zero on failure.
-allowed-tools: Read, Bash
----
-
-# /test — Run the Project's Test Suite
-
-The `/test` command is a thin wrapper around `bizar test-gate`. It
-auto-detects the project's test runner and runs the full suite.
-
-## What It Does
-
-1. Detects the project stack (one of: npm, pytest, cargo, go).
-2. Runs the matching test command:
-   - `package.json` present → `npm test`
-   - `pyproject.toml` present → `pytest`
-   - `Cargo.toml` present → `cargo test`
-   - `go.mod` present → `go test ./...`
-3. Streams the test output to the user.
-4. Exits non-zero on test failures.
-
-## Exit Codes
-
-| Code | Meaning                          |
-|------|----------------------------------|
-| 0    | All tests passed                 |
-| 1    | Tests failed                     |
-| 2    | No test runner detected          |
-
-## How to Use
-
-Invoke `/test` from the project root. The test gate inspects the
-current directory and picks the right runner.
-
-To run a specific Bizar test scope instead, pass arguments to
-`/test`. The arguments are forwarded to the underlying runner and
-are available as `$ARGUMENTS` (or `$1`, `$2`, ...):
-
-- `/test plugins/bizar` — run only the Bizar plugin tests (bun)
-- `/test --testPathPattern=auth` — pass a flag to the runner
-
-## When to Use
-
-- After any non-trivial code change
-- Before committing (`@hermod` should always run `/test` first)
-- As the final step of a `/team` mission — gate quality with tests
-
-## Common Patterns
-
-- **Post-implementation gate** — after Thor and Tyr finish their
-  parallel work, dispatch `thor` to run `/test`. If failures, fix
-  and re-run.
-- **PR check** — run `/test` in CI before merging.
-- **Debug a flaky test** — run `/test` twice; if the second run
-  passes, the test is order-dependent (fix or quarantine).
-
-## Related
-
-- `bizar test-gate` — the underlying CLI command
-- `bizar doctor` — checks the Bizar install itself (not project tests)
-- `make check` — typecheck + tests (full pipeline, dev only)
-- `make e2e` — Bizar plugin end-to-end (dev only)

package/.claude/worktrees/kanban-overhaul/.claude/commands/validate.md

@@ -1,68 +0,0 @@
----
-description: Validate the Bizar install end-to-end — settings.json parses, plugin entry resolves, all agent files are installed, all skills/rules/hooks/commands are in place, provider config is sane, and the 9router gateway is reachable.
-allowed-tools: Read, Bash
----
-
-# /validate — Validate the Bizar Install
-
-The `/validate` command runs `bizar doctor` with the full check
-battery. It is the fastest way to confirm the install is healthy
-and Claude Code will see every Bizar component on the next session
-start.
-
-## What It Checks
-
-1. **claude-reachable** — `claude --version` exits 0.
-2. **claude-config-valid** — `~/.claude/settings.json` parses as JSON.
-3. **plugin-entry-present** — `settings.json` has a Bizar plugin entry.
-4. **plugin-path-resolves** — the plugin path on disk actually exists.
-5. **agent-files-installed** — all agent files in `~/.claude/agents/`
-   (odin, vor, frigg, quick, mimir, heimdall, hermod, thor, baldr,
-   tyr, vidarr, forseti, semble-search, agent-browser).
-6. **tools-available** — at least one of `headroom`, `semble`, `skills`
-   is on PATH.
-7. **provider-config-sanity** — `provider.9router` (preferred) or
-   `provider.minimax` (legacy) has at least one sane model entry.
-8. **9router-reachable** — 9Router gateway responds at
-   `http://localhost:20128/api/health` (lenient — warn, don't fail).
-9. **slash-commands-installed** — all Bizar slash commands
-   (`/audit`, `/explain`, `/init`, `/learn`, `/plan`, `/plow-through`,
-   `/pr-review`, `/tailscale-serve`, `/visual-plan`, `/bizar`,
-   `/team`, `/test`, `/validate`) are present in `~/.claude/commands/`.
-10. **skills-installed** — every subdir of `config/skills/` is
-    mirrored to `~/.claude/skills/`.
-11. **rules-installed** — every `.md` in `config/rules/` is mirrored
-    to `~/.claude/rules/`.
-12. **hooks-installed** — `config/hooks/` is mirrored to
-    `~/.claude/hooks/`.
-
-## How to Use
-
-Invoke `/validate` (no arguments) from anywhere. It is read-only and
-never modifies the system.
-
-If any check fails, the user should run `bizar update` to refresh the
-install, or `bizar repair` to fix common issues.
-
-## Output
-
-Each check prints ✓ or ✗ with a one-line message. The summary line
-shows the total pass/fail count. Exits non-zero if any check fails.
-
-For machine-readable output, the user can run `bizar doctor --json`
-directly.
-
-## When to Use
-
-- After a manual config edit — confirm nothing broke.
-- Before a `/team` mission — confirm the install is healthy so the
-  team has all the tools it needs.
-- After `bizar install` or `bizar update` — the install already
-  calls doctor, but a manual re-run is cheap.
-- When a user asks "why is Claude Code misbehaving?" — start here.
-
-## Related
-
-- `bizar doctor` — the underlying CLI command
-- `bizar update` — refreshes the install (re-runs syncConfigExtras)
-- `bizar repair` — fixes common issues (stale symlinks, version drift)

package/.claude/worktrees/kanban-overhaul/.claude/commands/visual-plan.md

@@ -1,24 +0,0 @@
----
-description: Toggle the Bizar visual plan canvas or check its current status.
----
-
-# /visual-plan — Toggle Visual Plan Canvas
-
-Toggle the Bizar visual plan canvas or check its current status.
-
-The full arguments are available as `$ARGUMENTS` (and `$1`).
-
-## Usage
-
-```
-/visual-plan on   — Enable the visual plan canvas in the current session
-/visual-plan off  — Disable the visual plan canvas
-/visual-plan status — Show whether the canvas is currently enabled
-```
-
-## Notes
-
-The visual plan canvas shows plans as interactive node graphs. When enabled, plan elements appear as draggable nodes on a canvas that agents can manipulate directly. This is useful for visual thinkers or when planning complex multi-step work.
-
-This is a session-level toggle — it affects the current Claude Code
-session only.

package/.claude/worktrees/kanban-overhaul/.claude/hooks/README.md

@@ -1,92 +0,0 @@
-# Bizar Harness — Claude Code Hooks
-
-This directory contains the **Claude Code** adapter scripts that back the
-project-scoped hooks defined in `.claude/settings.json`.
-
-They replaced the legacy Cline hooks in `config/hooks/*` (PreToolUse,
-PostToolUse, TaskStart, TaskResume, UserPromptSubmit). The Claude Code
-set is the only supported path on Bizar v6.3.0+; the legacy Cline hooks
-are no longer installed.
-
-## Adapter inventory
-
-| File | Event | Matcher | Behaviour |
-|------|-------|---------|-----------|
-| `pretooluse-editwrite.mjs`   | `PreToolUse`       | `Write\|Edit\|MultiEdit\|Bash` | Blocks `.env`/secrets/lockfiles; warns on `console.log`/`debugger`/`.only()` in `src/`; scans `Bash` against the 36-pattern dangerous-pattern list |
-| `posttooluse-editwrite.mjs`  | `PostToolUse`      | `Edit\|Write\|MultiEdit`        | Logs edit/write latency + size to `~/.config/bizar/hook-logs/` |
-| `sessionstart-prime.mjs`     | `SessionStart`     | (none — branches on `source`)  | Primes the agent with CLAUDE.md, AGENTS.md, feature_list state |
-| `userpromptsubmit-tag.mjs`   | `UserPromptSubmit` | (none)                          | Tags prompt for `/team`, `/plow-through`, `/test`, `/validate`, `/plan`, `/audit`, `/pr-review` |
-| `sessionend-recall.mjs`      | `SessionEnd`       | (none)                          | Records a session summary to the memory vault at `~/.bizar_home/memory/projects/<name>/sessions/` |
-
-## I/O contract
-
-Claude Code invokes each hook as a child process, pipes JSON on
-**stdin**, and reads JSON on **stdout**. The five adapters above all
-read using `process.stdin` and write a single JSON object to
-`process.stdout`.
-
-### Stdin (varies by event)
-
-```jsonc
-// PreToolUse
-{
-  "session_id": "...",
-  "transcript_path": "...",
-  "cwd": "/home/drb0rk/Projects/BizarHarness",
-  "hook_event_name": "PreToolUse",
-  "tool_name": "Write",
-  "tool_input": { "file_path": "/path", "content": "..." }
-}
-
-// PostToolUse  (adds tool_response)
-// UserPromptSubmit  (uses .user_prompt instead of .tool_name/.tool_input)
-// SessionStart  (uses .source = "startup" | "resume" | "clear")
-// SessionEnd  (uses .reason = "exit" | "clear" | "logout" | "prompt_input_exit")
-```
-
-### Stdout
-
-```jsonc
-// Block (PreToolUse only)
-{
-  "hookSpecificOutput": {
-    "hookEventName": "PreToolUse",
-    "permissionDecision": "deny",
-    "permissionDecisionReason": "..."
-  }
-}
-
-// Inject context for the next turn (all events)
-{
-  "hookSpecificOutput": {
-    "hookEventName": "PostToolUse",   // or PreToolUse / SessionStart / UserPromptSubmit
-    "additionalContext": "..."
-  }
-}
-```
-
-For a non-blocking `PreToolUse`, omitting `permissionDecision` and
-emitting only `additionalContext` is a soft hint to the model — the
-tool call still proceeds.
-
-## Claude Code lifecycle mapping
-
-The legacy Cline plugin used four separate events (PreToolUse,
-PostToolUse, TaskStart, TaskResume, UserPromptSubmit). Claude Code
-collapses "task start" and "task resume" into a single `SessionStart`
-event that branches on `source` (`startup`, `resume`, `clear`). The
-adapter `sessionstart-prime.mjs` reads `source` and applies the
-appropriate priming logic for each branch.
-
-## Disabling an individual hook
-
-Easiest path: comment out the matching `hooks` array entry in
-`.claude/settings.json`. Removing the executable bit
-(`chmod -x …`) also works — Claude Code silently skips hooks it can't
-execute.
-
-## See also
-
-- `cli/provision-claude.mjs` — installs Claude Code hooks to
-  `~/.claude/hooks/` and wires them via `~/.claude/settings.json`.
-- `.claude/settings.json` — the project hook wiring

package/.claude/worktrees/kanban-overhaul/.claude/hooks/posttooluse-editwrite.mjs

@@ -1,91 +0,0 @@
-#!/usr/bin/env node
-// PostToolUse — Bizar harness hook (Claude Code format).
-//
-// Runs AFTER a tool call against `Edit`, `Write`, or `MultiEdit`
-// completes. Logs the edit/write to `~/.config/bizar/hook-logs/` so
-// downstream observability tooling can aggregate latency + size.
-//
-// Claude Code stdin shape:
-//   {
-//     "session_id": "...",
-//     "transcript_path": "...",
-//     "cwd": "...",
-//     "hook_event_name": "PostToolUse",
-//     "tool_name": "Edit",
-//     "tool_input": { "file_path": "..." },
-//     "tool_response": { ... }
-//   }
-//
-// Claude Code stdout shape:
-//   { "hookSpecificOutput": {
-//       "hookEventName": "PostToolUse",
-//       "additionalContext": "note for the next turn"
-//   }}
-//
-// Behaviour:
-//   1. Append-only JSONL log under ~/.config/bizar/hook-logs/.
-//   2. Remind to run /test after edits to src/ files.
-
-'use strict';
-
-import fs from 'node:fs';
-import os from 'node:os';
-import path from 'node:path';
-
-let raw = '';
-process.stdin.setEncoding('utf8');
-process.stdin.on('data', (chunk) => { raw += chunk; });
-process.stdin.on('end', () => {
-  let input = {};
-  try { input = JSON.parse(raw || '{}'); } catch { input = {}; }
-
-  const toolName = String(input.tool_name || '');
-  const toolInput =
-    (input.tool_input && typeof input.tool_input === 'object') ? input.tool_input : {};
-  const filePath = String(toolInput.file_path || '');
-  const sessionId = String(input.session_id || '');
-
-  // 1. Append-only JSONL log under ~/.config/bizar/hook-logs/.
-  try {
-    const logDir = path.join(os.homedir(), '.config', 'bizar', 'hook-logs');
-    fs.mkdirSync(logDir, { recursive: true });
-    const today = new Date().toISOString().slice(0, 10);
-    const logFile = path.join(logDir, `post-tool-use-${today}.jsonl`);
-    const line = JSON.stringify({
-      ts: new Date().toISOString(),
-      tool: toolName,
-      success: true, // PostToolUse only fires on success in Claude Code
-      executionTimeMs: 0, // not exposed in Claude Code hook input
-      sessionId: sessionId || null,
-    });
-    fs.appendFileSync(logFile, line + '\n');
-  } catch { /* best-effort */ }
-
-  const notes = [];
-
-  // 2. Skip the slow-tool warning — Claude Code doesn't surface tool
-  //    latency in hook input, so no signal to warn on.
-  //
-  // 3. Surface a reminder for Write/Edit/MultiEdit on src/ paths.
-  if (
-    /^(write|edit|multiedit)$/i.test(toolName) &&
-    /\/(src|plugins|packages)\//.test(filePath.toLowerCase())
-  ) {
-    notes.push(
-      "Bizar PostToolUse: if you're done editing, run `/test` to gate the change with tests before claiming complete.",
-    );
-  }
-
-  if (notes.length === 0) {
-    process.stdout.write('{}\n');
-    return;
-  }
-
-  const out = {
-    hookSpecificOutput: {
-      hookEventName: 'PostToolUse',
-      additionalContext: notes.join(' '),
-    },
-  };
-  process.stdout.write(JSON.stringify(out) + '\n');
-});

package/.claude/worktrees/kanban-overhaul/.claude/hooks/pretooluse-bash.mjs

@@ -1,81 +0,0 @@
-#!/usr/bin/env node
-/**
- * pretooluse-bash.mjs — Claude Code PreToolUse hook (matcher: Bash).
- *
- * Runs the Bizar dangerous-pattern scanner on Bash commands before they
- * execute. Blocks destructive patterns (rm -rf, sudo, SSRF, etc.) via
- * the 36-pattern DANGEROUS_PATTERNS regex list maintained in
- * `@polderlabs/bizar-sdk/dist/dangerous-patterns.js`.
- *
- * Claude Code PreToolUse input schema:
- *   { session_id, cwd, hook_event_name: "PreToolUse", tool_name: "Bash", tool_input: { command, description } }
- *
- * Claude Code PreToolUse output schema:
- *   { hookSpecificOutput: { hookEventName, permissionDecision, permissionDecisionReason }, additionalContext? }
- */
-
-'use strict';
-
-let raw = '';
-process.stdin.setEncoding('utf8');
-process.stdin.on('data', (chunk) => { raw += chunk; });
-process.stdin.on('end', () => {
-  let input = {};
-  try { input = JSON.parse(raw); } catch { input = {}; }
-  const toolInput = (input.tool_input && typeof input.tool_input === 'object') ? input.tool_input : {};
-  const command = String(toolInput.command || '');
-
-  // Inline dangerous-pattern scanner. Mirrors
-  // `packages/sdk/src/dangerous-patterns.ts` — kept inline so this hook
-  // works without importing the SDK at runtime.
-  const DANGEROUS = [
-    { name: 'rm-rf-root', pattern: /\brm\s+(-\w*r\w*f\w*\s+)*\/\s*$/i, decision: 'deny', reason: 'Recursive delete of root filesystem' },
-    { name: 'rm-rf-etc', pattern: /\brm\s+(-\w*r\w*f\w*\s+)*\/(etc|var|usr|boot|home)\b/i, decision: 'deny', reason: 'Recursive delete of system directory' },
-    { name: 'rm-rf-wildcard', pattern: /\brm\s+-\w*r\w*f\w*\s+\*/i, decision: 'ask', reason: 'Recursive delete with wildcard' },
-    { name: 'rm-rf-home', pattern: /\brm\s+(-\w*r\w*f\w*\s+)*~?\//i, decision: 'ask', reason: 'Recursive delete of home directory' },
-    { name: 'mkfs', pattern: /\bmkfs(\.\w+)?\s+\/dev\//i, decision: 'deny', reason: 'Format filesystem' },
-    { name: 'dd-of-dev', pattern: /\bdd\s+.*of=\/dev\//i, decision: 'deny', reason: 'dd to raw device' },
-    { name: 'sudo', pattern: /(^|\s|;|&&|\|\|)sudo\b/i, decision: 'ask', reason: 'Sudo escalation' },
-    { name: 'su-root', pattern: /\bsu\s+-?\s*root\b/i, decision: 'ask', reason: 'su to root' },
-    { name: 'curl-metadata', pattern: /169\.254\.169\.254/i, decision: 'deny', reason: 'AWS metadata IP' },
-    { name: 'curl-google-metadata', pattern: /metadata\.google\.internal/i, decision: 'deny', reason: 'GCP metadata hostname' },
-    { name: 'curl-azure-metadata', pattern: /metadata\.azure\.com/i, decision: 'deny', reason: 'Azure metadata hostname' },
-    { name: 'nc-backdoor', pattern: /\bnc\s+(-[a-z]*\s+)*-[a-z]*e\b/i, decision: 'deny', reason: 'nc execute backdoor' },
-    { name: 'wget-pipe-shell', pattern: /\bwget\s+.*\|\s*(ba)?sh\b/i, decision: 'deny', reason: 'wget piped to shell' },
-    { name: 'curl-pipe-shell', pattern: /\bcurl\s+.*\|\s*(ba)?sh\b/i, decision: 'deny', reason: 'curl piped to shell' },
-    { name: 'kill-init', pattern: /\bkill\s+(-9\s+)?1\b/i, decision: 'deny', reason: 'Kill PID 1 (init)' },
-    { name: 'shutdown', pattern: /\b(shutdown|halt|poweroff|reboot)\b/i, decision: 'deny', reason: 'System shutdown' },
-    { name: 'xmrig', pattern: /\b(xmrig|minerd|cpuminer|cgminer)\b/i, decision: 'deny', reason: 'Crypto miner' },
-    { name: 'read-shadow', pattern: /\/etc\/(shadow|passwd|sudoers)\b/i, decision: 'ask', reason: 'Read sensitive system file' },
-    { name: 'read-ssh', pattern: /\.ssh\//i, decision: 'deny', reason: 'Read SSH keys' },
-    { name: 'read-aws-creds', pattern: /~\/\.aws\/credentials/i, decision: 'deny', reason: 'Read AWS credentials' },
-    { name: 'path-traversal', pattern: /(\.\.\/){2,}/i, decision: 'ask', reason: 'Multiple path traversal' },
-    { name: 'git-force-push-main', pattern: /\bgit\s+push\s+(-f|--force)(\s+--.*)?\s+origin\s+(main|master)\b/i, decision: 'deny', reason: 'Force push to main/master' },
-    { name: 'git-reset-hard', pattern: /\bgit\s+reset\s+--hard\b/i, decision: 'ask', reason: 'git reset --hard' },
-  ];
-
-  for (const p of DANGEROUS) {
-    if (p.pattern.test(command)) {
-      const decision = p.decision === 'deny' ? 'deny' : 'ask';
-      const out = {
-        hookSpecificOutput: {
-          hookEventName: 'PreToolUse',
-          permissionDecision: decision,
-          permissionDecisionReason: `dangerous_pattern:${p.name}: ${p.reason}`,
-        },
-      };
-      process.stdout.write(JSON.stringify(out) + '\n');
-      return;
-    }
-  }
-
-  // No dangerous pattern — allow.
-  const out = {
-    hookSpecificOutput: {
-      hookEventName: 'PreToolUse',
-      permissionDecision: 'allow',
-    },
-    additionalContext: `Bizar PreToolUse(Bash): scanned ${command.length} chars, no dangerous pattern.`,
-  };
-  process.stdout.write(JSON.stringify(out) + '\n');
-});