@codyswann/lisa 1.0.0 → 1.0.5

package/all/copy-overwrite/.claude/REFERENCE.md

@@ -0,0 +1,519 @@
+# Claude Code Hooks & Settings Reference
+
+Complete expert reference for Claude Code hooks and `.claude/settings.json` configuration.
+
+---
+
+## Hook Event Types (13 total)
+
+| Event | Trigger | Matcher Support |
+|-------|---------|-----------------|
+| `SessionStart` | Session begins/resumes | `startup`, `resume`, `clear`, `compact` |
+| `SessionEnd` | Session terminates | `clear`, `logout`, `prompt_input_exit`, `other` |
+| `Setup` | `--init`, `--init-only`, `--maintenance` flags | `init`, `maintenance` |
+| `PreToolUse` | Before tool call processes | Tool names (regex supported) |
+| `PostToolUse` | After tool succeeds | Tool names (regex supported) |
+| `PostToolUseFailure` | After tool fails | Tool names (regex supported) |
+| `PermissionRequest` | Permission dialog appears | Tool names |
+| `SubagentStart` | Subagent spawns | N/A |
+| `SubagentStop` | Subagent finishes | N/A |
+| `Stop` | Main agent finishes (not interrupted) | N/A |
+| `UserPromptSubmit` | User submits prompt | N/A |
+| `Notification` | Notifications sent | `permission_prompt`, `idle_prompt`, `auth_success`, `elicitation_dialog` |
+| `PreCompact` | Before context compaction | `manual`, `auto` |
+
+---
+
+## Complete settings.json Schema
+
+```json
+{
+  "permissions": {
+    "allow": ["Bash(npm run:*)", "Read"],
+    "ask": ["Bash(git push:*)"],
+    "deny": ["Bash(rm -rf:*)", "Read(./.env.*)"],
+    "additionalDirectories": ["../docs/"],
+    "defaultMode": "default | acceptEdits | plan | bypassPermissions | dontAsk",
+    "disableBypassPermissionsMode": "disable | warn"
+  },
+
+  "hooks": {},
+  "disableAllHooks": false,
+  "allowManagedHooksOnly": false,
+
+  "env": {
+    "ANTHROPIC_API_KEY": "sk-...",
+    "BASH_DEFAULT_TIMEOUT_MS": "30000",
+    "BASH_MAX_TIMEOUT_MS": "7200000",
+    "CLAUDE_CODE_ENABLE_TELEMETRY": "1",
+    "CLAUDE_CODE_MAX_OUTPUT_TOKENS": "32000",
+    "MAX_THINKING_TOKENS": "31999",
+    "OTEL_METRICS_EXPORTER": "otlp",
+    "OTEL_EXPORTER_OTLP_ENDPOINT": "...",
+    "CUSTOM_VAR": "value"
+  },
+
+  "model": "string",
+  "alwaysThinkingEnabled": false,
+
+  "statusLine": { "type": "command", "command": "<HOME>/.claude/statusline.sh" },
+  "outputStyle": "string",
+  "language": "string",
+  "spinnerTipsEnabled": true,
+  "terminalProgressBarEnabled": true,
+  "showTurnDuration": true,
+
+  "attribution": {
+    "commit": "🤖 Generated with Claude Code\n\nCo-Authored-By: ...",
+    "pr": "🤖 Generated with Claude Code"
+  },
+  "includeCoAuthoredBy": false,
+
+  "forceLoginMethod": "claudeai | console",
+  "forceLoginOrgUUID": "string",
+  "apiKeyHelper": "string",
+  "awsAuthRefresh": "string",
+  "awsCredentialExport": "string",
+  "otelHeadersHelper": "string",
+
+  "enableAllProjectMcpServers": false,
+  "enabledMcpjsonServers": [],
+  "disabledMcpjsonServers": [],
+  "allowedMcpServers": [],
+  "deniedMcpServers": [],
+
+  "enabledPlugins": { "plugin-name@marketplace": true },
+  "extraKnownMarketplaces": {},
+  "strictKnownMarketplaces": [],
+
+  "respectGitignore": true,
+  "fileSuggestion": { "type": "command", "command": "..." },
+  "additionalDirectories": [],
+
+  "sandbox": {
+    "enabled": true,
+    "autoAllowBashIfSandboxed": true,
+    "excludedCommands": ["git", "docker"],
+    "allowUnsandboxedCommands": true,
+    "network": {
+      "allowUnixSockets": ["<HOME>/.ssh/agent-socket"],
+      "allowLocalBinding": true,
+      "httpProxyPort": 8080,
+      "socksProxyPort": 8081
+    },
+    "enableWeakerNestedSandbox": false
+  },
+
+  "cleanupPeriodDays": 30,
+  "autoUpdatesChannel": "stable | latest",
+  "plansDirectory": "string",
+  "companyAnnouncements": []
+}
+```
+
+---
+
+## Hook Configuration Structure
+
+```json
+{
+  "hooks": {
+    "EventName": [
+      {
+        "matcher": "ToolPattern",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "bash-cmd",
+            "timeout": 60
+          }
+        ]
+      }
+    ]
+  }
+}
+```
+
+### Hook Types
+
+| Type | Description |
+|------|-------------|
+| `command` | Executes bash command synchronously |
+| `prompt` | Sends JSON input to LLM (Haiku) for decision |
+
+**Prompt-based hooks** only support: `Stop`, `SubagentStop`, `UserPromptSubmit`, `PreToolUse`, `PermissionRequest`
+
+Prompt hooks must return JSON: `{"ok": true|false, "reason": "explanation"}`
+
+---
+
+## Environment Variables Available to Hooks
+
+| Variable | Description | Availability |
+|----------|-------------|--------------|
+| `CLAUDE_PROJECT_DIR` | Absolute path to project root | Always |
+| `CLAUDE_CODE_REMOTE` | `"true"` if web environment | Always |
+| `CLAUDE_ENV_FILE` | File path to persist env vars | SessionStart/Setup only |
+| `CLAUDE_PLUGIN_ROOT` | Plugin directory path | Plugin hooks only |
+
+---
+
+## Exit Code Behavior
+
+| Exit Code | Behavior |
+|-----------|----------|
+| **0** | Success - stdout processed for JSON or added as context |
+| **2** | Blocking error - prevents action, stderr shown to Claude |
+| **Other** | Non-blocking error - stderr shown in verbose mode |
+
+---
+
+## Blocking vs Non-Blocking Hooks
+
+### Blocking Hooks (Exit Code 2)
+
+| Event | Effect |
+|-------|--------|
+| `PreToolUse` | Blocks tool call, shows stderr to Claude |
+| `PermissionRequest` | Denies permission, shows stderr to Claude |
+| `PostToolUse` | Shows stderr to Claude (tool already ran) |
+| `UserPromptSubmit` | Erases submitted prompt, shows stderr to user only |
+| `Stop`/`SubagentStop` | Blocks stoppage, shows stderr to Claude |
+| `Setup`/`SessionStart`/`SessionEnd`/`PreCompact`/`Notification` | Shows stderr to user only |
+
+---
+
+## Matcher Patterns
+
+| Pattern | Matches |
+|---------|---------|
+| `Write` | Exact match only |
+| `Edit\|Write` | Either tool (regex OR) |
+| `npm run:*` | Prefix + word boundary (`npm run lint` yes, `npm runtest` no) |
+| `ls*` | Glob anywhere (`ls -la`, `lsof`) |
+| `Notebook.*` | Regex pattern (all Notebook tools) |
+| `mcp__server__.*` | All MCP server tools |
+| `""` or omit | All tools/events |
+
+### Common Tool Names for Matching
+
+**Core Tools:**
+`Bash`, `Write`, `Edit`, `Read`, `Glob`, `Grep`, `WebFetch`, `WebSearch`
+
+**Notebook Tools:**
+`Notebook`, `NotebookEdit`
+
+**Agent & Task Tools:**
+`Task`, `TaskCreate`, `TaskUpdate`, `TaskList`, `TaskGet`
+
+**Skill & Planning Tools:**
+`Skill`, `EnterPlanMode`, `ExitPlanMode`, `AskUserQuestion`
+
+**Advanced Tools:**
+`LSP`, `Computer`, `ToolSearch`, `TeammateTool`
+
+**MCP Tools Pattern:**
+`mcp__<server>__<tool>` (e.g., `mcp__memory__create_entities`)
+
+---
+
+## Settings File Locations & Precedence
+
+| Scope | Location | Shared | Precedence |
+|-------|----------|--------|------------|
+| Managed | System-level `managed-settings.json` | Yes (IT deployed) | Highest (cannot override) |
+| User | `<HOME>/.claude/settings.json` | No | Normal |
+| Project | `.claude/settings.json` | Yes (in git) | Normal |
+| Local | `.claude/settings.local.json` | No (gitignored) | Normal |
+
+### Managed Settings Locations
+
+| OS | Path |
+|----|------|
+| macOS | `<SYSTEM_CONFIG>/ClaudeCode/` |
+| Linux/WSL | `<SYSTEM_CONFIG>/claude-code/` |
+| Windows | `<SYSTEM_CONFIG>\ClaudeCode\` |
+
+---
+
+## Hook Output Handling by Event
+
+| Event | stdout | stderr |
+|-------|--------|--------|
+| PreToolUse/PostToolUse/Stop/SubagentStop | Shown in verbose mode, JSON parsed for control | Shown in verbose mode |
+| Notification/SessionEnd | Logged to debug only (`--debug`) | Logged to debug only |
+| UserPromptSubmit/SessionStart/Setup | **Added to Claude's context** | Shown to user in verbose mode |
+
+---
+
+## Hook Input Schema (Common Fields)
+
+All hook inputs include:
+
+```json
+{
+  "session_id": "abc123",
+  "transcript_path": "/path/to/transcript.jsonl",
+  "cwd": "/current/working/directory",
+  "permission_mode": "default | plan | acceptEdits | dontAsk | bypassPermissions",
+  "hook_event_name": "EventName"
+}
+```
+
+### PreToolUse Input
+
+```json
+{
+  "tool_name": "Bash | Write | Edit | Read | Glob | Grep | WebFetch | WebSearch | Task",
+  "tool_input": {
+    "command": "...",
+    "description": "...",
+    "timeout": 120000,
+    "run_in_background": false
+  },
+  "tool_use_id": "toolu_01ABC123..."
+}
+```
+
+### PostToolUse Input
+
+```json
+{
+  "tool_name": "...",
+  "tool_input": {},
+  "tool_response": {},
+  "tool_use_id": "toolu_01ABC123..."
+}
+```
+
+### Notification Input
+
+```json
+{
+  "message": "Claude needs your permission to use Bash",
+  "notification_type": "permission_prompt | idle_prompt | auth_success | elicitation_dialog"
+}
+```
+
+### UserPromptSubmit Input
+
+```json
+{
+  "prompt": "User's prompt text"
+}
+```
+
+### Stop Input
+
+```json
+{
+  "stop_hook_active": false
+}
+```
+
+### SubagentStop Input
+
+```json
+{
+  "stop_hook_active": false,
+  "agent_id": "def456",
+  "agent_transcript_path": "/path/to/subagent/transcript.jsonl"
+}
+```
+
+### SubagentStart Input
+
+```json
+{
+  "agent_id": "agent-abc123",
+  "agent_type": "Explore | Bash | Plan | CustomAgentName"
+}
+```
+
+### SessionStart Input
+
+```json
+{
+  "source": "startup | resume | clear | compact",
+  "model": "claude-sonnet-4-20250514",
+  "agent_type": "AgentName"
+}
+```
+
+### SessionEnd Input
+
+```json
+{
+  "reason": "exit | clear | logout | prompt_input_exit | other"
+}
+```
+
+### Setup Input
+
+```json
+{
+  "trigger": "init | maintenance"
+}
+```
+
+### PreCompact Input
+
+```json
+{
+  "trigger": "manual | auto",
+  "custom_instructions": ""
+}
+```
+
+---
+
+## Hook-Specific Output Schemas
+
+### PreToolUse Output
+
+```json
+{
+  "hookSpecificOutput": {
+    "hookEventName": "PreToolUse",
+    "permissionDecision": "allow | deny | ask",
+    "permissionDecisionReason": "Optional explanation",
+    "updatedInput": {
+      "field_name": "new_value"
+    },
+    "additionalContext": "Additional context for Claude"
+  }
+}
+```
+
+### PermissionRequest Output
+
+```json
+{
+  "hookSpecificOutput": {
+    "hookEventName": "PermissionRequest",
+    "decision": {
+      "behavior": "allow | deny",
+      "message": "Why denied",
+      "interrupt": true,
+      "updatedInput": {}
+    }
+  }
+}
+```
+
+### PostToolUse Output
+
+```json
+{
+  "hookSpecificOutput": {
+    "hookEventName": "PostToolUse",
+    "additionalContext": "Feedback for Claude"
+  },
+  "decision": "block | undefined",
+  "reason": "Why blocked (if decision=block)"
+}
+```
+
+### UserPromptSubmit Output
+
+```json
+{
+  "hookSpecificOutput": {
+    "hookEventName": "UserPromptSubmit",
+    "additionalContext": "Additional context to add"
+  },
+  "decision": "block | undefined",
+  "reason": "Why blocked"
+}
+```
+
+### Stop/SubagentStop Output
+
+```json
+{
+  "decision": "block | undefined",
+  "reason": "Why blocked (required when blocked)"
+}
+```
+
+### SessionStart/Setup Output
+
+```json
+{
+  "hookSpecificOutput": {
+    "hookEventName": "SessionStart | Setup",
+    "additionalContext": "Context to inject"
+  }
+}
+```
+
+---
+
+## Permission Evaluation Order
+
+First match wins:
+
+1. **Deny rules** (highest priority)
+2. **Ask rules**
+3. **Allow rules** (lowest priority)
+
+---
+
+## Wildcard Patterns in Permissions
+
+| Pattern | Description |
+|---------|-------------|
+| `:*` | Prefix matching with word boundary |
+| `*` | Glob matching anywhere |
+| `**` | Recursive glob |
+
+Examples:
+- `Bash(npm run:*)` - matches "npm run lint" but NOT "npm runtest"
+- `Bash(ls*)` - matches both "ls -la" and "lsof"
+- `Read(./secrets/**)` - matches all files under ./secrets/
+
+---
+
+## Best Practices
+
+1. **Blocking behavior for enforcement**: When creating hooks for linting, code quality, or static analysis, use blocking behavior (exit 2) so Claude receives feedback and can fix errors.
+
+2. **Notification hooks**: Notification-only hooks should exit 0 since they don't require Claude to take action.
+
+3. **JSON parsing**: Never parse JSON in shell scripts using grep/sed/cut/awk—always use `jq` for robust JSON handling.
+
+4. **Environment persistence**: Use `CLAUDE_ENV_FILE` in SessionStart/Setup hooks to persist environment variables across subsequent bash commands:
+   ```bash
+   echo 'export VAR=value' >> "$CLAUDE_ENV_FILE"
+   ```
+
+5. **Timeout configuration**: Default is 60 seconds per hook. Configure per-hook with `"timeout"` field (in seconds).
+
+---
+
+---
+
+## Claude Code System Prompts Reference
+
+Comprehensive catalog of all 110+ system prompts used by Claude Code v2.1.19+. Maintained by Piebald AI and updated within minutes of each Claude Code release.
+
+**What It Contains:**
+- Main system prompt and ~40 system reminders
+- 18+ tool descriptions (Bash, Write, Edit, Read, WebFetch, WebSearch, etc.)
+- Specialized agent prompts (Explore, Plan, Task execution)
+- Creation assistants (Agent architect, CLAUDE.md generator, status line setup)
+- Slash command implementations (/security-review, /review-pr, /pr-comments)
+- Utility functions (summarization, session management, security analysis)
+
+**Why It's Useful:**
+Understand exactly how Claude Code is instructed to operate across different scenarios. Reference for customization via tweakcc tool.
+
+**Repository:** [Piebald-AI/claude-code-system-prompts](https://github.com/Piebald-AI/claude-code-system-prompts)
+
+---
+
+## Sources
+
+- [Claude Code Hooks Reference](https://docs.anthropic.com/en/docs/claude-code/hooks)
+- [Claude Code Settings Documentation](https://docs.anthropic.com/en/docs/claude-code/settings)
+- [Claude Code System Prompts Repository](https://github.com/Piebald-AI/claude-code-system-prompts)

package/all/copy-overwrite/.claude/agents/skill-evaluator.md

@@ -80,13 +80,13 @@ Examples:
 #### Criterion 5: Non-Redundancy (Does it already exist?)
 
 **PASS**: Not already covered by existing skills or project documentation
-**FAIL**: Duplicates existing skills or belongs in CLAUDE.md/PROJECT_RULES.md
+**FAIL**: Duplicates existing skills or belongs in CLAUDE.md/.claude/rules/PROJECT_RULES.md
 
 Before recommending skill creation, always check:
 
 1. Existing skills in `.claude/skills/`
 2. CLAUDE.md for project-level instructions
-3. PROJECT_RULES.md for project rules
+3. .claude/rules/PROJECT_RULES.md for project rules
 
 ### Step 3: Make the Decision
 
@@ -101,7 +101,7 @@ Before recommending skill creation, always check:
 **DO NOT CREATE SKILL** if any criterion fails. Instead:
 
 - If **redundant** → **OMIT ENTIRELY** - reference the existing documentation instead
-- If too simple → Suggest adding to CLAUDE.md or PROJECT_RULES.md
+- If too simple → Suggest adding to CLAUDE.md or .claude/rules/PROJECT_RULES.md
 - If too narrow → Just apply the knowledge directly, no documentation needed
 - If temporary → Document as a code comment or temporary note
 
@@ -111,7 +111,7 @@ Before evaluating other criteria, check if the content already exists:
 
 1. **Existing skill covers it** → OMIT - just invoke the existing skill
 2. **Already in CLAUDE.md** → OMIT - it's already enforced
-3. **Already in PROJECT_RULES.md** → OMIT - it's already documented
+3. **Already in .claude/rules/PROJECT_RULES.md** → OMIT - it's already documented
 
 If redundant, stop evaluation and report: "This is already covered by [source]. No action needed."
 
@@ -190,9 +190,9 @@ When evaluating content, provide this assessment:
 - Reusability: ✅ Ongoing rule
 - Complexity: ❌ Single rule, easily remembered
 - Stability: ✅ Established pattern
-- Non-Redundancy: ✅ Not in PROJECT_RULES.md
+- Non-Redundancy: ✅ Not in .claude/rules/PROJECT_RULES.md
 
-**Decision**: ADD TO RULES → Add to PROJECT_RULES.md instead
+**Decision**: ADD TO RULES → Add to .claude/rules/PROJECT_RULES.md instead
 
 ### Example 4: Should OMIT ENTIRELY (Already Covered)
 
@@ -227,7 +227,7 @@ When evaluating content, provide this assessment:
 1. **Check for redundancy FIRST** - Before any other evaluation, verify it's not already covered
 2. **Default to NOT creating skills** - Skills should be rare
 3. **OMIT is a valid decision** - If it exists elsewhere, don't duplicate it
-4. **Simple rules go in CLAUDE.md or PROJECT_RULES.md** - Not every rule needs a skill
+4. **Simple rules go in CLAUDE.md or .claude/rules/PROJECT_RULES.md** - Not every rule needs a skill
 5. **When in doubt, don't create a skill** - It's better to have fewer, high-quality skills
 6. **Skills are for Claude, not humans** - Focus on what would help another Claude instance
 

package/all/copy-overwrite/.claude/agents/test-coverage-agent.md

@@ -0,0 +1,17 @@
+---
+name: test-coverage-agent
+description: Adds comprehensive test coverage to source files. Use for writing meaningful unit and integration tests.
+tools: Read, Write, Edit, Bash, Grep, Glob
+model: opus
+auto-invoke: true
+---
+
+You are a test coverage specialist. When given a file:
+
+1. Analyze the source file to understand its functionality
+2. Identify untested code paths, edge cases, and error conditions
+3. Write comprehensive, meaningful tests (not just coverage padding)
+4. Follow the project's existing test patterns and conventions
+5. Ensure tests are readable and maintainable
+
+Focus on testing behavior, not implementation details.

package/all/copy-overwrite/.claude/commands/git/commit.md

@@ -18,9 +18,10 @@ Create conventional commits for current changes. Optional hint: $ARGUMENTS
 ### Apply these requirements
 
 1. **Branch Check**: If on `dev`, `staging`, or `main`, create a feature branch named after the changes
-2. **Commit Strategy**: Group related changes into conventional commits (feat, fix, chore, docs, etc.)
-3. **Commit Creation**: Stage and commit changes in logical batches with clear messages
-4. **Verification**: Ensure all changes are committed
+2. **Commit Strategy**: Group related changes into logical conventional commits (feat, fix, chore, docs, etc.)
+3. **Commit ALL Files**: Every file must be assigned to a commit group - no file gets left out or unstaged
+4. **Commit Creation**: Stage and commit each group with clear messages
+5. **Verification**: Run `git status` to confirm working directory is clean - must show "nothing to commit"
 
 ### Use conventional commit format
 
@@ -34,10 +35,13 @@ Create conventional commits for current changes. Optional hint: $ARGUMENTS
 
 ### Never
 
-- use `--no-verify` or `--no-verify` flags
+- use `--no-verify` flag
 - attempt to bypass tests or quality checks
 - skip tests or quality checks
-- stash changes
+- stash changes - ALL changes must be committed
+- skip or exclude any files from the commit
+- leave uncommitted changes in the working directory
+- ask the user which files to commit - commit everything
 
 ## Execute
 

package/all/copy-overwrite/.claude/commands/git/submit-pr.md

@@ -25,7 +25,7 @@ Push current branch and create or update a pull request. Optional hint: $ARGUMEN
 
 1. **Branch Check**: Verify not on `dev`, `staging`, or `main` (cannot create PR from protected branches)
 2. **Commit Check**: Ensure all changes are committed before pushing
-3. **Push**: Push current branch to remote with `-u` flag
+3. **Push**: Push current branch to remote with `-u` flag and the following environment variable - GIT_SSH_COMMAND="ssh -o ServerAliveInterval=30 -o ServerAliveCountMax=5"
 4. **PR Management**:
    - Check for existing PR on this branch
    - If exists: Update description with latest changes