agentic-orchestrator 0.1.28 → 0.2.0

package/PROMPTS.md

@@ -0,0 +1,161 @@
+# Cortex TMS: AI Prompt Library
+
+This file contains project-aware prompt templates for streamlined AI collaboration. Use `cortex-tms prompt <name>` to quickly access these prompts.
+
+---
+
+## init-session
+
+Review `NEXT-TASKS.md`, `docs/core/ARCHITECTURE.md`, and `CLAUDE.md`. Summarize current priorities and propose a step-by-step plan for the next task.
+
+---
+
+## feature
+
+Implement `[FEATURE]`. First, check `ARCHITECTURE.md` for constraints, follow `PATTERNS.md` for style, and validate logic against `DOMAIN-LOGIC.md`.
+
+---
+
+## debug
+
+I'm seeing `[ISSUE]`. Check `docs/core/TROUBLESHOOTING.md` for known issues, then trace the logic through `ARCHITECTURE.md` to find the failure point.
+
+---
+
+## review
+
+Review the current changes against `PATTERNS.md`. Flag any violations of project conventions and suggest specific fixes.
+
+---
+
+## refactor
+
+Refactor `[COMPONENT]`. Improve structure per `PATTERNS.md` while ensuring behavior remains strictly compliant with `DOMAIN-LOGIC.md`.
+
+---
+
+## decision
+
+We need to decide on `[TOPIC]`. Draft a new Architecture Decision Record in `docs/core/DECISIONS.md` using the project's standard format.
+
+---
+
+## finish
+
+Task complete. Execute the Maintenance Protocol:
+
+1. Run `pnpm run docs:sync` to ensure version integrity across documentation
+2. Update `NEXT-TASKS.md` with task status and next priorities
+3. Run `node bin/cortex-tms.js validate --strict` to verify project health
+4. Follow Git Protocol: commit with conventional format and co-authorship
+5. Suggest next priority from backlog
+
+---
+
+## bootstrap
+
+You have just been added to a project that uses Cortex TMS for AI governance.
+The documentation files exist but contain placeholder content that needs to be
+populated with real project information.
+
+**Your task**: Analyze this codebase and populate the TMS documentation files
+as DRAFTS for human review.
+
+**Step 1 - Understand the project**:
+
+- Read `package.json` (or equivalent) for tech stack and dependencies
+- Scan the directory structure to understand components
+- Read entry points (`src/index.*`, `src/app.*`, `bin/*`)
+- Read any existing README.md for project context
+
+**Step 2 - Populate ARCHITECTURE.md** (`docs/core/ARCHITECTURE.md`):
+
+- Fill "Quick Context" with what the project does, who it's for, key constraint
+- Fill "System Overview" with 2-3 sentences about the system
+- Fill "Component Map" table with real components from the codebase
+- Fill "Core Data Flow" with the actual happy path
+- Mark all populated sections with: `<!-- AI-DRAFT: Review before treating as canonical -->`
+
+**Step 3 - Customize CLAUDE.md**:
+
+- Update CLI commands with the project's actual commands (from package.json scripts)
+- Add project-specific operational loop steps
+- Add any project-specific rules or conventions you observed
+
+**Step 4 - Customize copilot-instructions.md** (`.github/copilot-instructions.md`):
+
+- Fill in the actual tech stack
+- Add real conventions observed in the code
+- Add project-specific prohibitions
+
+**Step 5 - Populate NEXT-TASKS.md**:
+
+- If there are open TODOs, issues, or incomplete features, suggest initial tasks
+- Otherwise, suggest "Review AI-generated documentation" as first task
+
+**Rules**:
+
+- Be SPECIFIC to this project - no generic advice
+- Reference actual file paths from the codebase
+- Mark every generated section with `<!-- AI-DRAFT -->` comment
+- Ask the user to confirm before writing each file
+- If unsure about something, ASK rather than guess
+- Do NOT open or send content from `.env*`, `*.pem`, `id_rsa*`, or similar secret files
+- Skip `node_modules/`, `dist/`, `build/`, `coverage/`, `.git/`
+
+---
+
+## populate-architecture
+
+Analyze this codebase and populate `docs/core/ARCHITECTURE.md`:
+
+1. Read `package.json` for tech stack, then scan `src/` directory structure
+2. Fill "Quick Context": what it does, who it's for, key constraint
+3. Fill "Component Map": real components with responsibilities and tech stack
+4. Fill "Core Data Flow": trace the main happy path from entry point to output
+5. Fill "Deployment & Infrastructure" from config files and CI workflows
+
+Be specific to THIS project. Reference actual file paths. Mark sections with
+`<!-- AI-DRAFT: Review before treating as canonical -->`.
+Present the draft and ask for confirmation before writing.
+
+---
+
+## discover-patterns
+
+Review this codebase and document patterns for `docs/core/PATTERNS.md`:
+
+1. Scan representative files in `src/` for recurring conventions
+2. Identify: naming conventions, error handling, component structure, testing patterns
+3. For each pattern, find a canonical example (actual file path + code snippet)
+4. Document as "Pattern N: Name" with Rule, Anti-Pattern, and Canonical Example
+
+Extract ACTUAL patterns from code, not generic best practices.
+Mark all entries with `<!-- AI-DRAFT -->`. Present draft for review before writing.
+
+---
+
+## extract-rules
+
+Extract domain-specific rules for `docs/core/DOMAIN-LOGIC.md`:
+
+1. Look for validation logic (schemas, validators, form rules)
+2. Look for business rules (calculations, state machines, conditionals)
+3. Look for integration constraints (API limits, timeouts, rate limits)
+4. Look for security rules (auth checks, input sanitization)
+
+For each rule: write a clear imperative statement, cite the source file,
+and mark severity (critical/important/guideline).
+Mark all entries with `<!-- AI-DRAFT -->`. Present draft for review before writing.
+
+---
+
+## Usage Tips
+
+- Replace placeholders like `[FEATURE]`, `[ISSUE]`, `[COMPONENT]`, `[TOPIC]` with your specific context
+- Customize these prompts to match your team's vocabulary and workflow
+- Run `cortex-tms prompt --list` to see all available prompts
+
+<!-- @cortex-tms-version 4.0.2 -->
+
+<!-- @cortex-tms-version 4.0.2 -->

package/README.md

@@ -4,13 +4,13 @@
 [![Nx Workspace](https://img.shields.io/badge/Nx-Monorepo-143055.svg)](https://nx.dev/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-A deterministic, MCP-first control plane for parallel multi-agent feature development.
+A semi-deterministic, MCP-first control plane for parallel multi-agent feature development.
 
 It combines:
 
 - a **deterministic kernel** (state, locks, collisions, gates, evidence)
 - a **supervisor runtime** (orchestrator/planner/builder/qa session coordination)
-- a **CLI contract** (`aop <run|status|resume|stop|delete|cleanup|rollback|init|dashboard|retry|send|attach>`) with deterministic spec ingestion, runtime control, monitoring, and cleanup
+- a **CLI contract** (`aop <run|status|resume|stop|delete|cleanup|rollback|init|dashboard|retry|merge|send|attach|completion>`) with deterministic spec ingestion, runtime control, monitoring, cleanup, and shell registration output
 
 > For complete normative requirements, see [agentic_orchestrator_spec.md](spec-files/agentic_orchestrator_spec.md).
 > For implementation continuity and the latest status, see [progress.md](spec-files/progress.md).
@@ -216,7 +216,26 @@ All commands print JSON to stdout. Errors use:
 { "ok": false, "error": { "code": "...", "message": "...", "details": { ... } } }
 ```
 
-Available commands: `run`, `status`, `resume`, `stop`, `delete`, `cleanup`, `rollback`, `init`, `dashboard`, `retry`, `send`, `attach`, `help`.
+Available commands: `run`, `status`, `resume`, `stop`, `delete`, `cleanup`, `rollback`, `init`, `dashboard`, `retry`, `merge`, `send`, `attach`, `completion`, `help`.
+
+### `completion`
+
+Prints a shell registration script for Bash, Zsh, or Fish. This command writes raw shell code to stdout so it can be evaluated or sourced directly.
+
+Examples:
+
+```bash
+# Bash
+eval "$(aop completion bash)"
+
+# Zsh
+eval "$(aop completion zsh)"
+
+# Fish
+aop completion fish | source
+```
+
+The hidden backend `aop __complete` is reserved for the generated shell scripts and always degrades gracefully to empty output on malformed input.
 
 ### `run`
 
@@ -503,32 +522,32 @@ aop help dashboard
 
 Supported options:
 
-| Option                            | Description                                                                       |
-| --------------------------------- | --------------------------------------------------------------------------------- | ------------------------------------------------------------ | ------------------------------------------ | ---------- | --------- | ------------------ |
-| `-fi <PATH>`                      | Run exactly one spec file                                                         |
-| `-fl <PATH>`                      | Resolve all `**/*.md` files recursively in a folder (deterministic lexical order) |
-| `--batch`                         | Batch-ingest all specs; skip already-active features (`run`)                      |
-| `--agent-provider <codex\\        | claude\\                                                                          | gemini\\                                                     | custom\\                                   | kiro-cli\\ | copilot>` | Provider selection |
-| `--agent-model <model-id>`        | Model selection                                                                   |
-| `--agent-config <json-object>`    | Additional provider-specific agent config (for example command/args payload)      |
-| `--provider-config-env <ENV_VAR>` | Provider auth/config env var name for API-backed providers                        |
-| `--transport <mcp\\               | inprocess>`                                                                       | Tool transport selection (default `mcp`)                     |
-| `--takeover-stale-run`            | Allow stale run-lease takeover during `run`                                       |
-| `--project <name>`                | Select project from `multi-project.yaml` (run, status, resume, retry)             |
-| `--feature-id <id>`               | Target feature for `delete`, `cleanup`, `rollback`, `retry`, `send`, `attach`     |
-| `--checkpoint <id>`               | Target checkpoint id for `rollback`                                               |
-| `--dry-run [true\\                | false]`                                                                           | Preview-only mode for `delete`, `cleanup`, and `rollback`    |
-| `--yes`                           | Confirm destructive `delete` or `cleanup` execution                               |
-| `--remove-worktree [true\\        | false]`                                                                           | Whether to remove `.worktrees/<feature_id>` (default `true`) |
-| `--remove-branch <none\\          | safe\\                                                                            | force>`                                                      | Local branch cleanup mode (default `none`) |
-| `--summary`                       | Condensed one-line-per-feature output (`status`)                                  |
-| `--all`                           | Aggregate across all projects (`status`); delete all features (`delete`)          |
-| `--auto`                          | Non-interactive init with all defaults (`init`)                                   |
-| `--force`                         | Overwrite existing config (`init`); force retry regardless of state (`retry`)     |
-| `--port <number>`                 | Dashboard listen port (`aop dashboard`) or stop target port (`aop stop`)          |
-| `--dev`                           | Start dashboard in Next.js dev mode                                               |
-| `--foreground`                    | Keep dashboard server in the foreground                                           |
-| `--message <text>`                | Message to deliver to agent session (`send`)                                      |
+| Option                                                                | Description                                                                       |
+| --------------------------------------------------------------------- | --------------------------------------------------------------------------------- |
+| `-fi <PATH>`                                                          | Run exactly one spec file                                                         |
+| `-fl <PATH>`                                                          | Resolve all `**/*.md` files recursively in a folder (deterministic lexical order) |
+| `--batch`                                                             | Batch-ingest all specs; skip already-active features (`run`)                      |
+| `--agent-provider <codex\|claude\|gemini\|custom\|kiro-cli\|copilot>` | Provider selection                                                                |
+| `--agent-model <model-id>`                                            | Model selection                                                                   |
+| `--agent-config <json-object>`                                        | Additional provider-specific agent config (for example command/args payload)      |
+| `--provider-config-env <ENV_VAR>`                                     | Provider auth/config env var name for API-backed providers                        |
+| `--transport <mcp\|inprocess>`                                        | Tool transport selection (default `mcp`)                                          |
+| `--takeover-stale-run`                                                | Allow stale run-lease takeover during `run`                                       |
+| `--project <name>`                                                    | Select project from `multi-project.yaml` (run, status, resume, retry)             |
+| `--feature-id <id>`                                                   | Target feature for `delete`, `cleanup`, `rollback`, `retry`, `send`, `attach`     |
+| `--checkpoint <id>`                                                   | Target checkpoint id for `rollback`                                               |
+| `--dry-run [true\|false]`                                             | Preview-only mode for `delete`, `cleanup`, and `rollback`                         |
+| `--yes`                                                               | Confirm destructive `delete` or `cleanup` execution                               |
+| `--remove-worktree [true\|false]`                                     | Whether to remove `.worktrees/<feature_id>` (default `true`)                      |
+| `--remove-branch <none\|safe\|force>`                                 | Local branch cleanup mode (default `none`)                                        |
+| `--summary`                                                           | Condensed one-line-per-feature output (`status`)                                  |
+| `--all`                                                               | Aggregate across all projects (`status`); delete all features (`delete`)          |
+| `--auto`                                                              | Non-interactive init with all defaults (`init`)                                   |
+| `--force`                                                             | Overwrite existing config (`init`); force retry regardless of state (`retry`)     |
+| `--port <number>`                                                     | Dashboard listen port (`aop dashboard`) or stop target port (`aop stop`)          |
+| `--dev`                                                               | Start dashboard in Next.js dev mode                                               |
+| `--foreground`                                                        | Keep dashboard server in the foreground                                           |
+| `--message <text>`                                                    | Message to deliver to agent session (`send`)                                      |
 
 Validation behavior:
 
@@ -609,6 +628,7 @@ Coverage parser:
 - role-specific system prompt paths
 - default provider/model values with optional `runtime.provider_config_env` for API-backed providers
 - optional `runtime.provider_configs.<provider>` objects for provider-specific payloads (for example `kiro-cli chat --agent dev`)
+- optional `runtime.role_provider_overrides.<role>` to pin a smaller/larger model per phase (planner/builder/qa) without changing global defaults
 - `runtime.execution_mode` default (`deterministic|interactive`) plus `runtime.interactive.*` checkpoint/watchdog controls (`checkpoint_interval_ms`, `watchdog_poll_interval_ms`, `max_uncommitted_changes`, `validation_on_checkpoint`, `revert_on_violation`, `violation_severity`)
 - `worktree.post_create` commands and `worktree.symlinks` for workspace hook automation
 - stack-specific examples: [`example-configurations/node/`](example-configurations/node) and [`example-configurations/java/`](example-configurations/java)
@@ -627,6 +647,19 @@ runtime:
     violation_severity: warning
 ```
 
+Per-phase model override example:
+
+```yaml
+runtime:
+  role_provider_overrides:
+    planner:
+      model: gpt-4o-mini
+    builder:
+      model: gpt-4o
+    qa:
+      model: gpt-4o-mini
+```
+
 ### Multi-Project (`config/agentic/orchestrator/multi-project.yaml`)
 
 Optional. Defines multiple independent projects in one repository. Pass `--project <name>` on the CLI to scope commands to a project. `aop status --all` aggregates across all projects. Each project can override `policy` and `gates` paths.
@@ -705,6 +738,8 @@ For feature `my_feature`:
 
 Base branch used for cut is configured in `policy.yaml` (`worktree.base_branch`, currently `main`).
 
+`repo.reconcile_mainline` reconciles against the committed base-branch ref, not against uncommitted changes in your local working tree. If you want local fixes to be consumed automatically by reconciliation, commit them onto the base branch first or apply them directly in the feature worktree.
+
 ### Manual review workflow (recommended)
 
 1. Inspect working changes in the feature worktree:
@@ -823,6 +858,68 @@ The MCP contracts validator checks: protocol pinning, `catalog.json`/`tools.md`/
 
 ---
 
+## IntentGraph dependency index
+
+This repository maintains an automatically generated dependency and navigation index produced by **[IntentGraph](https://github.com/Raytracer76/IntentGraph)**. The index exists to help AI agents and contributors understand file relationships, dependency chains, and reverse-dependency paths **without repeatedly scanning the repository by hand**.
+
+### Why this exists
+
+For large or evolving codebases, repeatedly searching the repo to answer questions like:
+
+- What depends on this file or module?
+- What imports or calls this?
+- What is the dependency path from A to B?
+- What files are likely impacted by this change?
+
+...is slow, noisy, and wasteful. The IntentGraph index provides a machine-readable relationship map that should be consulted first for dependency and navigation questions.
+
+### Standard usage policy
+
+Use the IntentGraph index as the **authoritative first-stop source** for repository relationship discovery.
+
+- Consult the index first for dependency-chain and impact-analysis questions.
+- Open source files only after the index has narrowed the candidate set.
+- Do **not** manually recurse through large portions of the repo just to discover relationships that already exist in the generated index.
+
+### Index location
+
+- `./docs/core/intentgraph.index.json`
+
+### Install IntentGraph locally
+
+Install IntentGraph locally before attempting to generate or refresh the index.
+
+```bash
+pip install intentgraph
+```
+
+If your local environment uses a virtual environment, install it inside that environment.
+
+### Generate or refresh the index
+
+Regenerate the index whenever you:
+
+- Add, remove, move, or rename files
+- Change imports, exports, module boundaries, or dependency wiring
+- Refactor public APIs or shared interfaces
+- Make architecture changes that alter call paths or dependency relationships
+
+Example command:
+
+```bash
+intentgraph analyze . --output ./docs/core/intentgraph.index.json --level full
+```
+
+### Important usage note
+
+Do not paste the entire generated JSON into prompts, comments, or pull request descriptions. Query or inspect only the smallest relevant subset needed for the current task. The point of the index is to reduce unnecessary repo traversal and reduce context waste.
+
+### Staleness policy
+
+If the generated index conflicts with the current repository contents, treat the index as stale. Regenerate it before relying on it for dependency analysis.
+
+---
+
 ## Current Status
 
 From [progress.md](spec-files/progress.md):

package/agentic/orchestrator/agents.yaml

@@ -2,15 +2,16 @@ version: 1
 roles:
   planner:
     system_prompt_path: agentic/orchestrator/prompts/planner.system.md
+    intake_system_prompt_path: agentic/orchestrator/prompts/planner-intake.system.md
   builder:
     system_prompt_path: agentic/orchestrator/prompts/builder.system.md
   qa:
     system_prompt_path: agentic/orchestrator/prompts/qa.system.md
 missing_prompt_behavior: ignore
 runtime:
-  default_provider: custom
-  default_model: local-default
-  execution_mode: deterministic
+  default_provider: claude
+  default_model: claude-sonnet-4-6
+  execution_mode: interactive
   interactive:
     watchdog_poll_interval_ms: 2000
     checkpoint_interval_ms: 30000

package/agentic/orchestrator/defaults/policy.defaults.yaml

@@ -88,6 +88,7 @@ merge_policy:
     - merge_commit
     - squash
     - rebase
+  merge_target: base_branch
 worktree:
   base_branch: main
 rbac:
@@ -98,6 +99,11 @@ rbac:
     - feature.state_get
     - feature.state_patch
     - feature.log_append
+    - feature.question_create
+    - feature.question_list
+    - feature.question_answer
+    - feature.send_message
+    - feature.intake_submit
     - plan.submit
     - plan.get
     - plan.update
@@ -108,6 +114,7 @@ rbac:
     - repo.read_file
     - repo.search
     - repo.diff_bundle
+    - repo.reconcile_mainline
     - gates.list
     - gates.run
     - evidence.latest
@@ -120,15 +127,18 @@ rbac:
     - report.feature_summary
     - feature.ready_to_merge
     - feature.delete
-    - feature.send_message
     - cost.record
     - cost.get
     - performance.record_outcome
     - performance.get_analytics
+    - replay.timeline_get
   planner:
     - feature.get_context
     - feature.state_get
+    - feature.question_create
+    - feature.intake_submit
     - plan.submit
+    - plan.get
     - plan.update
     - collisions.scan
     - repo.read_file
@@ -137,6 +147,7 @@ rbac:
   builder:
     - feature.get_context
     - feature.state_get
+    - feature.question_create
     - repo.apply_patch
     - repo.status
     - repo.diff
@@ -166,8 +177,13 @@ supervisor:
   max_parallel_gate_runs: 2
   agent_idle_threshold_ms: 300000
 cleanup:
-  auto_after_merge: false
-  grace_period_seconds: 3600
+  auto_after_merge:
+    worktree:
+      enabled: false
+      grace_period_seconds: 3600
+    aop:
+      enabled: false
+      grace_period_seconds: 3600
 dashboard:
   enabled: true
   port: 3000
@@ -223,3 +239,23 @@ issue_tracker:
   type: github
   config:
     repo: ""
+innovation:
+  replay_debugger: true
+  tamper_evident_evidence: false
+observability:
+  raw_agent_logs_enabled: false
+  raw_agent_logs_retention_days: 7
+  replay:
+    enabled: true
+    write_mode: best_effort
+    retention:
+      max_runs: 100
+      max_entries_per_run: 50000
+  evidence_integrity:
+    enabled: false
+    verify_on_read: false
+    algorithm_version: 1
+diagnostics:
+  doctor:
+    enabled: true
+    strict_fails_on: critical

package/agentic/orchestrator/gates.yaml

@@ -9,12 +9,24 @@ profiles:
   default:
     modes:
       fast:
+        - name: typecheck
+          cmd: ["npm", "run", "typecheck"]
+          timeout_seconds: 600
         - name: lint
-          cmd: ["npm", "run", "nx", "--", "run", "control-plane:test"]
+          cmd: ["npm", "run", "lint"]
           timeout_seconds: 600
       full:
-        - name: test_with_coverage
-          cmd: ["npm", "run", "nx", "--", "run", "control-plane:test"]
+        - name: test
+          cmd: ["npm", "run", "test"]
+          timeout_seconds: 1200
+        - name: lint
+          cmd: ["npm", "run", "lint"]
+          timeout_seconds: 1200
+        - name: typecheck
+          cmd: ["npm", "run", "typecheck"]
+          timeout_seconds: 1200
+        - name: validate_all
+          cmd: ["npm", "run", "validate:all"]
           timeout_seconds: 1200
       merge:
         - name: merge_safety

package/agentic/orchestrator/policy.yaml

@@ -71,6 +71,20 @@ execution:
   env_allowlist:
     - PATH
     - HOME
+  human_input:
+    enabled: true
+    timeout_ms: 900000
+    on_timeout: block_feature
+    max_open_questions_per_feature: 1
+    context_answer_history_limit: 3
+planning:
+  intake:
+    enabled: true
+    require_verified_manifest_before_planning: true
+    allow_auto_promotion: false
+    auto_promotion_min_confidence: high
+    max_intake_question_rounds: 2
+    on_unresolved_ambiguity: awaiting_input
 implementation:
   workspace: nx
 testing:
@@ -103,10 +117,15 @@ supervisor:
   max_parallel_gate_runs: 2
   agent_idle_threshold_ms: 300000
 cleanup:
-  auto_after_merge: false
-  grace_period_seconds: 3600
+  auto_after_merge:
+    worktree:
+      enabled: false
+      grace_period_seconds: 3600
+    aop:
+      enabled: false
+      grace_period_seconds: 3600
 observability:
-  raw_agent_logs_enabled: false
+  raw_agent_logs_enabled: true
   raw_agent_logs_retention_days: 60
 notifications:
   enabled: false
@@ -142,6 +161,7 @@ rbac:
     - repo.read_file
     - repo.search
     - repo.diff_bundle
+    - repo.reconcile_mainline
     - gates.list
     - gates.run
     - evidence.latest
@@ -155,19 +175,27 @@ rbac:
     - feature.ready_to_merge
     - feature.delete
     - feature.send_message
+    - feature.question_create
+    - feature.question_list
+    - feature.question_answer
+    - feature.intake_submit
     - cost.record
     - cost.get
     - performance.record_outcome
     - performance.get_analytics
+    - replay.timeline_get
   planner:
     - feature.get_context
     - feature.state_get
     - plan.submit
+    - plan.get
     - plan.update
+    - feature.intake_submit
     - collisions.scan
     - repo.read_file
     - repo.search
     - performance.get_analytics
+    - feature.question_create
   builder:
     - feature.get_context
     - feature.state_get
@@ -178,6 +206,7 @@ rbac:
     - repo.search
     - gates.run
     - performance.get_analytics
+    - feature.question_create
   qa:
     - feature.get_context
     - feature.state_get
@@ -190,6 +219,21 @@ rbac:
     - qa.test_index_get
     - qa.test_index_update
     - performance.get_analytics
+    - feature.question_create
+  reconciler:
+    - feature.get_context
+    - feature.state_get
+    - feature.log_append
+    - plan.get
+    - plan.update
+    - repo.apply_patch
+    - repo.status
+    - repo.diff
+    - repo.read_file
+    - repo.search
+    - gates.run
+    - performance.get_analytics
+    - feature.question_list
   system:
     - "*"