Add orchestrator training, channels, LiteLLM engine, and simplify learning taxonomy

Major changes across parallel sessions:

- Add orchestrator SFT & GRPO training subpackage (learning/orchestrator/)
  with episode types, multi-objective reward, prompt registry, policy model,
  RL environment, and registered learning policies
- Add structured THOUGHT/TOOL/INPUT/FINAL_ANSWER mode to OrchestratorAgent
- Add 15 channel backends (Discord, Slack, Telegram, Email, Webhook, IRC,
  Matrix, Teams, WhatsApp, Signal, Mattermost, BlueBubbles, Feishu,
  Google Chat, Webchat) with channel tools and config
- Add LiteLLM engine backend for unified LLM provider access
- Add RLM agent and REPL tool
- Remove ToolLearningPolicy — learning taxonomy now only targets
  Intelligence (LM weights/routing) and Agents (logic/ICL/tool strategies)
- Rename SFTPolicy to SFTRouterPolicy (backward-compat alias kept)
- Remove OpenClaw agent infrastructure (openclaw*.py, openclaw_bridge.py)
- Fix async streaming tests (asyncio.run vs deprecated get_event_loop)
- Fix server channel route tests (pytest.importorskip for optional fastapi)
- Track uv.lock for reproducibility
- Update CLAUDE.md and docs to reflect all changes

1676 tests pass, 37 skipped.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

docs/api/learning.md

@@ -39,8 +39,7 @@ The learning system defines a hierarchy of learning policy ABCs:
 
 - **`LearningPolicy`** -- base ABC for all learning policies
 - **`IntelligenceLearningPolicy`** -- specialization for model routing decisions
-- **`AgentLearningPolicy`** -- specialization for agent behavior advice
-- **`ToolLearningPolicy`** -- specialization for tool selection/configuration
+- **`AgentLearningPolicy`** -- specialization for agent behavior advice (ICL examples, tool-use strategies)
 
 ---
 
@@ -59,9 +58,9 @@ The learning system defines a hierarchy of learning policy ABCs:
     options:
       show_source: true
 
-### SFTPolicy
+### SFTRouterPolicy
 
-::: openjarvis.learning.sft_policy.SFTPolicy
+::: openjarvis.learning.sft_policy.SFTRouterPolicy
     options:
       show_source: true
       members_order: source

docs/architecture/intelligence.md

@@ -216,6 +216,6 @@ ctx = build_routing_context("```python\ndef hello():\n    pass\n```")
 
 ## Integration with Learning
 
-The `HeuristicRouter` implements the `RouterPolicy` ABC (now defined in `intelligence/_stubs.py`), which means it can be swapped out for a `TraceDrivenPolicy`, `SFTPolicy`, or any other policy via the `RouterPolicyRegistry`. See the [Learning & Traces](learning.md) documentation for details on how trace-driven routing and the broader `LearningPolicy` taxonomy work.
+The `HeuristicRouter` implements the `RouterPolicy` ABC (now defined in `intelligence/_stubs.py`), which means it can be swapped out for a `TraceDrivenPolicy`, `SFTRouterPolicy`, or any other policy via the `RouterPolicyRegistry`. See the [Learning & Traces](learning.md) documentation for details on how trace-driven routing and the broader `LearningPolicy` taxonomy work.
 
 The router is registered as `"heuristic"` in the `RouterPolicyRegistry` and is the default routing policy. Users can switch policies via the `--router` CLI flag or the `learning.default_policy` config setting.

docs/architecture/learning.md

@@ -6,13 +6,12 @@ The Learning system is a **cross-cutting concern** that connects all four pillar
 
 ## LearningPolicy ABC Taxonomy
 
-The learning system defines a hierarchy of learning policy ABCs. The base `LearningPolicy` ABC is specialized into three sub-ABCs corresponding to the three learnable concerns:
+The learning system defines a hierarchy of learning policy ABCs. The base `LearningPolicy` ABC is specialized into two sub-ABCs corresponding to the two learnable concerns:
 
 | ABC | Concern | Description |
 |-----|---------|-------------|
 | `IntelligenceLearningPolicy` | Model routing | Determines which model handles a query (replaces the legacy `RouterPolicy`) |
-| `AgentLearningPolicy` | Agent behavior | Advises on agent strategy (e.g., tool selection, turn limits) |
-| `ToolLearningPolicy` | Tool selection | Recommends tool configurations based on query characteristics |
+| `AgentLearningPolicy` | Agent behavior | Advises on agent strategy (e.g., ICL examples, tool selection, turn limits) |
 
 All learning policies are registered in the `LearningRegistry` (in `core/registry.py`).
 
@@ -66,14 +65,14 @@ The system ships with these router policies:
 | `heuristic` | `HeuristicRouter` | Active | Rule-based routing with 6 priority rules |
 | `learned` | `TraceDrivenPolicy` | Active | Learns from trace outcomes |
 | `grpo` | `GRPORouterPolicy` | Stub | Placeholder for future RL training |
-| `sft` | `SFTPolicy` | Active | Supervised fine-tuning policy (learns from labeled traces) |
+| `sft` | `SFTRouterPolicy` | Active | Trace-driven routing policy (learns query→model mapping); `SFTPolicy` is a backward-compat alias |
 
 And these additional learning policies (registered in `LearningRegistry`):
 
 | Registry Key | Policy Class | Taxonomy | Description |
 |-------------|-------------|----------|-------------|
 | `agent_advisor` | `AgentAdvisorPolicy` | `AgentLearningPolicy` | Advises on agent strategy based on trace patterns |
-| `icl_updater` | `ICLUpdaterPolicy` | `ToolLearningPolicy` | In-context learning updater for tool selection |
+| `icl_updater` | `ICLUpdaterPolicy` | `AgentLearningPolicy` | In-context learning updater — discovers ICL examples and multi-tool skills from traces |
 
 Users select a policy via `config.toml` or the `--router` CLI flag:
 
@@ -193,11 +192,13 @@ The online update uses a conservative strategy: it only switches the preferred m
 
 ---
 
-## SFTPolicy (Supervised Fine-Tuning)
+## SFTRouterPolicy (Trace-Driven Router)
 
-The `SFTPolicy` (in `learning/sft_policy.py`) is an `IntelligenceLearningPolicy` that learns routing decisions from labeled trace data using supervised fine-tuning principles. Unlike `TraceDrivenPolicy` which uses online aggregation, `SFTPolicy` trains a mapping from query features to model choices based on curated, high-quality trace examples.
+The `SFTRouterPolicy` (in `learning/sft_policy.py`) is an `IntelligenceLearningPolicy` that learns routing decisions from historical traces. It analyzes trace outcomes, groups by query class (code, math, short, long, general), and builds a `query_class → model` mapping from the highest-scoring model per class. A backward-compatible alias `SFTPolicy = SFTRouterPolicy` is provided for code that used the old name.
 
 ```python
+from openjarvis.learning.sft_policy import SFTRouterPolicy
+# or via the backward-compat alias:
 from openjarvis.learning.sft_policy import SFTPolicy
 ```
 
@@ -215,7 +216,7 @@ from openjarvis.learning.agent_advisor import AgentAdvisorPolicy
 
 ## ICLUpdaterPolicy
 
-The `ICLUpdaterPolicy` (in `learning/icl_updater.py`) is a `ToolLearningPolicy` that uses in-context learning to update tool selection and configuration. It analyzes recent tool-call traces to recommend which tools should be enabled for different query types.
+The `ICLUpdaterPolicy` (in `learning/icl_updater.py`) is an `AgentLearningPolicy` that uses in-context learning to discover reusable examples and multi-tool skill sequences from traces. It analyzes successful tool-call patterns to recommend ICL examples and skill libraries that update agent behavior.
 
 ```python
 from openjarvis.learning.icl_updater import ICLUpdaterPolicy