agentorch 2.0.2__tar.gz → 2.1.0__tar.gz

{agentorch-2.0.2 → agentorch-2.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentorch
-Version: 2.0.2
+Version: 2.1.0
 Summary: DevTorch — AI reasoning capture, audit trail, and governance for agent-assisted development
 Author-email: Hemant Joshi <hemant@flotorch.ai>
 License: Apache-2.0
@@ -33,6 +33,8 @@ Requires-Dist: fastapi>=0.115.0; extra == "proxy"
 Requires-Dist: uvicorn[standard]>=0.32.0; extra == "proxy"
 Requires-Dist: httpx>=0.27.0; extra == "proxy"
 Requires-Dist: pydantic>=2.9.0; extra == "proxy"
+Provides-Extra: embeddings
+Requires-Dist: sentence-transformers>=3.0.0; extra == "embeddings"
 Provides-Extra: mcp
 Requires-Dist: mcp>=1.0.0; extra == "mcp"
 Provides-Extra: cloud
@@ -46,16 +48,16 @@ Requires-Dist: azure-storage-blob>=12.19.0; extra == "cloud"
 Provides-Extra: broadcast
 Requires-Dist: watchdog>=3.0.0; extra == "broadcast"
 Provides-Extra: all
-Requires-Dist: agentorch[cloud,mcp,proxy,wrapper]; extra == "all"
+Requires-Dist: agentorch[cloud,embeddings,mcp,proxy,wrapper]; extra == "all"
 
 # DevTorch
 
 **Local-first AI agent governance.** DevTorch gives multi-agent systems a shared on-disk audit trail, coordination vector, privacy controls, and LLM integration layer — all without a remote service.
 
-> **Latest release:** `agentorch 2.0.2` — MCP bridge with automatic local fallback, plus upgrade/reinstall docs for all IDE integrations. Install or upgrade with `pip install -U agentorch`.
+> **Latest release:** `agentorch 2.1.0` — Reasoning Plus Learning: semantic relevance, learning deduplication, provenance tracking, LLM-based extraction, and a feedback loop. Install or upgrade with `pip install -U agentorch`.
 
 ```bash
-pip install -U agentorch>=2.0.2
+pip install -U agentorch>=2.1.0
 devtorch init
 devtorch setup          # connects to Claude Code, Cursor, or Antigravity
 devtorch debug timeline
@@ -120,7 +122,7 @@ DevTorch preserves your `.GCC/` store across upgrades. Newer releases add IDE co
 **Upgrade any time:**
 
 ```bash
-pip install -U agentorch>=2.0.2
+pip install -U agentorch>=2.1.0
 cd /path/to/your/project
 devtorch setup              # or: devtorch setup --target <ide>
 devtorch doctor
@@ -184,7 +186,7 @@ Open Claude Code in your project. On the first task Claude will call `devtorch_c
 
 **What you get:** Full RACP integration — reasoning commits, sensitivity events, coordination vector Θ, divergence detection, HITL conflict resolution.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup` will rewrite `.claude/hooks.json`, `.claude/settings.json`, and `CLAUDE.md` with the latest versions.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup` will rewrite `.claude/hooks.json`, `.claude/settings.json`, and `CLAUDE.md` with the latest versions.
 
 ---
 
@@ -228,7 +230,7 @@ devtorch doctor
 
 **What you get:** MCP tools, VS Code sidebar, optional HTTP proxy for passive capture of all LLM calls (`devtorch proxy start`).
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup --target cursor` will regenerate `.cursor/mcp.json` and update `CLAUDE.md`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup --target cursor` will regenerate `.cursor/mcp.json` and update `CLAUDE.md`.
 
 ---
 
@@ -260,7 +262,7 @@ devtorch doctor
 
 **What you get:** Full RACP integration. The multi-agent nature of Antigravity makes Θ accumulation especially valuable — multiple agents share the same reasoning state through `devtorch_context`.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup --target antigravity` will regenerate the MCP config and update `AGENTS.md`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup --target antigravity` will regenerate the MCP config and update `AGENTS.md`.
 
 ---
 
@@ -298,7 +300,7 @@ Open the project in OpenCode. The DevTorch MCP tools will be listed, and the mod
 
 **What you get:** Cloud-hosted MCP server integration with automatic local fallback, automatic reasoning commits via `AGENTS.md` instructions, and a shared Θ state across sessions.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch cloud client-config opencode --org <org> --repo <repo> --api-key <api-key>` will regenerate `opencode.json` with the latest bridge and fallback settings. **Restart OpenCode** after updating the config.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch cloud client-config opencode --org <org> --repo <repo> --api-key <api-key>` will regenerate `opencode.json` with the latest bridge and fallback settings. **Restart OpenCode** after updating the config.
 
 ---
 
@@ -330,7 +332,7 @@ devtorch sensitivity add --source you --concept auth --confidence 0.9 --disclosu
 
 **What you get:** Sidebar panel (read-only), git-commit-level event capture, manual reasoning commits. Full capture requires switching to Claude Code, Cursor, Kiro, or Antigravity.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup` will reinstall the git hook and update the governance instructions.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup` will reinstall the git hook and update the governance instructions.
 
 ---
 
@@ -395,7 +397,7 @@ devtorch doctor
 
 **What you get:** Full RACP integration equivalent to Claude Code — MCP tools, steering-document governance, agent action hooks.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2`, then re-run the manual steps above and refresh your steering document with the latest `AGENTS.md` text.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0`, then re-run the manual steps above and refresh your steering document with the latest `AGENTS.md` text.
 
 ---
 
@@ -428,7 +430,7 @@ All `invoke_model()` and `converse()` calls are identical. DevTorch injects the
 devtorch doctor
 ```
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -462,7 +464,7 @@ client = AzureOpenAI(
 devtorch doctor
 ```
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -488,7 +490,7 @@ codex "Add error handling to the payment service"
 
 All Codex calls are intercepted by the proxy, which injects the RACP prefix and captures responses into `.GCC/`.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0`.
 
 ---
 
@@ -522,7 +524,7 @@ devtorch doctor
 
 No API key required. Fully local — RACP context and captures stay on-device.
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -567,7 +569,7 @@ Supported `<ide>` values: `opencode`, `claude`, `cursor`, `vscode`, `stdio`.
 
 > **Updating from an older version:** You do **not** need to redeploy the Cloudflare Worker for client-side changes. Just regenerate the IDE config on each machine after upgrading `agentorch`:
 > ```bash
-> pip install -U agentorch>=2.0.2
+> pip install -U agentorch>=2.1.0
 > export DEVTORCH_MCP_DOMAIN=devtorch-mcp.YOUR_ACCOUNT.workers.dev
 > /path/to/DevTorch/Implementation/deploy/cloudflare-worker/install-client-config.sh \
 >   <ide> <org> <repo> <api-key>

{agentorch-2.0.2 → agentorch-2.1.0}/README.md

@@ -2,10 +2,10 @@
 
 **Local-first AI agent governance.** DevTorch gives multi-agent systems a shared on-disk audit trail, coordination vector, privacy controls, and LLM integration layer — all without a remote service.
 
-> **Latest release:** `agentorch 2.0.2` — MCP bridge with automatic local fallback, plus upgrade/reinstall docs for all IDE integrations. Install or upgrade with `pip install -U agentorch`.
+> **Latest release:** `agentorch 2.1.0` — Reasoning Plus Learning: semantic relevance, learning deduplication, provenance tracking, LLM-based extraction, and a feedback loop. Install or upgrade with `pip install -U agentorch`.
 
 ```bash
-pip install -U agentorch>=2.0.2
+pip install -U agentorch>=2.1.0
 devtorch init
 devtorch setup          # connects to Claude Code, Cursor, or Antigravity
 devtorch debug timeline
@@ -70,7 +70,7 @@ DevTorch preserves your `.GCC/` store across upgrades. Newer releases add IDE co
 **Upgrade any time:**
 
 ```bash
-pip install -U agentorch>=2.0.2
+pip install -U agentorch>=2.1.0
 cd /path/to/your/project
 devtorch setup              # or: devtorch setup --target <ide>
 devtorch doctor
@@ -134,7 +134,7 @@ Open Claude Code in your project. On the first task Claude will call `devtorch_c
 
 **What you get:** Full RACP integration — reasoning commits, sensitivity events, coordination vector Θ, divergence detection, HITL conflict resolution.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup` will rewrite `.claude/hooks.json`, `.claude/settings.json`, and `CLAUDE.md` with the latest versions.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup` will rewrite `.claude/hooks.json`, `.claude/settings.json`, and `CLAUDE.md` with the latest versions.
 
 ---
 
@@ -178,7 +178,7 @@ devtorch doctor
 
 **What you get:** MCP tools, VS Code sidebar, optional HTTP proxy for passive capture of all LLM calls (`devtorch proxy start`).
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup --target cursor` will regenerate `.cursor/mcp.json` and update `CLAUDE.md`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup --target cursor` will regenerate `.cursor/mcp.json` and update `CLAUDE.md`.
 
 ---
 
@@ -210,7 +210,7 @@ devtorch doctor
 
 **What you get:** Full RACP integration. The multi-agent nature of Antigravity makes Θ accumulation especially valuable — multiple agents share the same reasoning state through `devtorch_context`.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup --target antigravity` will regenerate the MCP config and update `AGENTS.md`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup --target antigravity` will regenerate the MCP config and update `AGENTS.md`.
 
 ---
 
@@ -248,7 +248,7 @@ Open the project in OpenCode. The DevTorch MCP tools will be listed, and the mod
 
 **What you get:** Cloud-hosted MCP server integration with automatic local fallback, automatic reasoning commits via `AGENTS.md` instructions, and a shared Θ state across sessions.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch cloud client-config opencode --org <org> --repo <repo> --api-key <api-key>` will regenerate `opencode.json` with the latest bridge and fallback settings. **Restart OpenCode** after updating the config.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch cloud client-config opencode --org <org> --repo <repo> --api-key <api-key>` will regenerate `opencode.json` with the latest bridge and fallback settings. **Restart OpenCode** after updating the config.
 
 ---
 
@@ -280,7 +280,7 @@ devtorch sensitivity add --source you --concept auth --confidence 0.9 --disclosu
 
 **What you get:** Sidebar panel (read-only), git-commit-level event capture, manual reasoning commits. Full capture requires switching to Claude Code, Cursor, Kiro, or Antigravity.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup` will reinstall the git hook and update the governance instructions.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup` will reinstall the git hook and update the governance instructions.
 
 ---
 
@@ -345,7 +345,7 @@ devtorch doctor
 
 **What you get:** Full RACP integration equivalent to Claude Code — MCP tools, steering-document governance, agent action hooks.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2`, then re-run the manual steps above and refresh your steering document with the latest `AGENTS.md` text.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0`, then re-run the manual steps above and refresh your steering document with the latest `AGENTS.md` text.
 
 ---
 
@@ -378,7 +378,7 @@ All `invoke_model()` and `converse()` calls are identical. DevTorch injects the
 devtorch doctor
 ```
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -412,7 +412,7 @@ client = AzureOpenAI(
 devtorch doctor
 ```
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -438,7 +438,7 @@ codex "Add error handling to the payment service"
 
 All Codex calls are intercepted by the proxy, which injects the RACP prefix and captures responses into `.GCC/`.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0`.
 
 ---
 
@@ -472,7 +472,7 @@ devtorch doctor
 
 No API key required. Fully local — RACP context and captures stay on-device.
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -517,7 +517,7 @@ Supported `<ide>` values: `opencode`, `claude`, `cursor`, `vscode`, `stdio`.
 
 > **Updating from an older version:** You do **not** need to redeploy the Cloudflare Worker for client-side changes. Just regenerate the IDE config on each machine after upgrading `agentorch`:
 > ```bash
-> pip install -U agentorch>=2.0.2
+> pip install -U agentorch>=2.1.0
 > export DEVTORCH_MCP_DOMAIN=devtorch-mcp.YOUR_ACCOUNT.workers.dev
 > /path/to/DevTorch/Implementation/deploy/cloudflare-worker/install-client-config.sh \
 >   <ide> <org> <repo> <api-key>

{agentorch-2.0.2 → agentorch-2.1.0}/agentorch.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentorch
-Version: 2.0.2
+Version: 2.1.0
 Summary: DevTorch — AI reasoning capture, audit trail, and governance for agent-assisted development
 Author-email: Hemant Joshi <hemant@flotorch.ai>
 License: Apache-2.0
@@ -33,6 +33,8 @@ Requires-Dist: fastapi>=0.115.0; extra == "proxy"
 Requires-Dist: uvicorn[standard]>=0.32.0; extra == "proxy"
 Requires-Dist: httpx>=0.27.0; extra == "proxy"
 Requires-Dist: pydantic>=2.9.0; extra == "proxy"
+Provides-Extra: embeddings
+Requires-Dist: sentence-transformers>=3.0.0; extra == "embeddings"
 Provides-Extra: mcp
 Requires-Dist: mcp>=1.0.0; extra == "mcp"
 Provides-Extra: cloud
@@ -46,16 +48,16 @@ Requires-Dist: azure-storage-blob>=12.19.0; extra == "cloud"
 Provides-Extra: broadcast
 Requires-Dist: watchdog>=3.0.0; extra == "broadcast"
 Provides-Extra: all
-Requires-Dist: agentorch[cloud,mcp,proxy,wrapper]; extra == "all"
+Requires-Dist: agentorch[cloud,embeddings,mcp,proxy,wrapper]; extra == "all"
 
 # DevTorch
 
 **Local-first AI agent governance.** DevTorch gives multi-agent systems a shared on-disk audit trail, coordination vector, privacy controls, and LLM integration layer — all without a remote service.
 
-> **Latest release:** `agentorch 2.0.2` — MCP bridge with automatic local fallback, plus upgrade/reinstall docs for all IDE integrations. Install or upgrade with `pip install -U agentorch`.
+> **Latest release:** `agentorch 2.1.0` — Reasoning Plus Learning: semantic relevance, learning deduplication, provenance tracking, LLM-based extraction, and a feedback loop. Install or upgrade with `pip install -U agentorch`.
 
 ```bash
-pip install -U agentorch>=2.0.2
+pip install -U agentorch>=2.1.0
 devtorch init
 devtorch setup          # connects to Claude Code, Cursor, or Antigravity
 devtorch debug timeline
@@ -120,7 +122,7 @@ DevTorch preserves your `.GCC/` store across upgrades. Newer releases add IDE co
 **Upgrade any time:**
 
 ```bash
-pip install -U agentorch>=2.0.2
+pip install -U agentorch>=2.1.0
 cd /path/to/your/project
 devtorch setup              # or: devtorch setup --target <ide>
 devtorch doctor
@@ -184,7 +186,7 @@ Open Claude Code in your project. On the first task Claude will call `devtorch_c
 
 **What you get:** Full RACP integration — reasoning commits, sensitivity events, coordination vector Θ, divergence detection, HITL conflict resolution.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup` will rewrite `.claude/hooks.json`, `.claude/settings.json`, and `CLAUDE.md` with the latest versions.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup` will rewrite `.claude/hooks.json`, `.claude/settings.json`, and `CLAUDE.md` with the latest versions.
 
 ---
 
@@ -228,7 +230,7 @@ devtorch doctor
 
 **What you get:** MCP tools, VS Code sidebar, optional HTTP proxy for passive capture of all LLM calls (`devtorch proxy start`).
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup --target cursor` will regenerate `.cursor/mcp.json` and update `CLAUDE.md`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup --target cursor` will regenerate `.cursor/mcp.json` and update `CLAUDE.md`.
 
 ---
 
@@ -260,7 +262,7 @@ devtorch doctor
 
 **What you get:** Full RACP integration. The multi-agent nature of Antigravity makes Θ accumulation especially valuable — multiple agents share the same reasoning state through `devtorch_context`.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup --target antigravity` will regenerate the MCP config and update `AGENTS.md`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup --target antigravity` will regenerate the MCP config and update `AGENTS.md`.
 
 ---
 
@@ -298,7 +300,7 @@ Open the project in OpenCode. The DevTorch MCP tools will be listed, and the mod
 
 **What you get:** Cloud-hosted MCP server integration with automatic local fallback, automatic reasoning commits via `AGENTS.md` instructions, and a shared Θ state across sessions.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch cloud client-config opencode --org <org> --repo <repo> --api-key <api-key>` will regenerate `opencode.json` with the latest bridge and fallback settings. **Restart OpenCode** after updating the config.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch cloud client-config opencode --org <org> --repo <repo> --api-key <api-key>` will regenerate `opencode.json` with the latest bridge and fallback settings. **Restart OpenCode** after updating the config.
 
 ---
 
@@ -330,7 +332,7 @@ devtorch sensitivity add --source you --concept auth --confidence 0.9 --disclosu
 
 **What you get:** Sidebar panel (read-only), git-commit-level event capture, manual reasoning commits. Full capture requires switching to Claude Code, Cursor, Kiro, or Antigravity.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2 && devtorch setup` will reinstall the git hook and update the governance instructions.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0 && devtorch setup` will reinstall the git hook and update the governance instructions.
 
 ---
 
@@ -395,7 +397,7 @@ devtorch doctor
 
 **What you get:** Full RACP integration equivalent to Claude Code — MCP tools, steering-document governance, agent action hooks.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2`, then re-run the manual steps above and refresh your steering document with the latest `AGENTS.md` text.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0`, then re-run the manual steps above and refresh your steering document with the latest `AGENTS.md` text.
 
 ---
 
@@ -428,7 +430,7 @@ All `invoke_model()` and `converse()` calls are identical. DevTorch injects the
 devtorch doctor
 ```
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -462,7 +464,7 @@ client = AzureOpenAI(
 devtorch doctor
 ```
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -488,7 +490,7 @@ codex "Add error handling to the payment service"
 
 All Codex calls are intercepted by the proxy, which injects the RACP prefix and captures responses into `.GCC/`.
 
-> **Updating from an older version:** `pip install -U agentorch>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch>=2.1.0`.
 
 ---
 
@@ -522,7 +524,7 @@ devtorch doctor
 
 No API key required. Fully local — RACP context and captures stay on-device.
 
-> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.0.2`.
+> **Updating from an older version:** `pip install -U agentorch[wrapper]>=2.1.0`.
 
 ---
 
@@ -567,7 +569,7 @@ Supported `<ide>` values: `opencode`, `claude`, `cursor`, `vscode`, `stdio`.
 
 > **Updating from an older version:** You do **not** need to redeploy the Cloudflare Worker for client-side changes. Just regenerate the IDE config on each machine after upgrading `agentorch`:
 > ```bash
-> pip install -U agentorch>=2.0.2
+> pip install -U agentorch>=2.1.0
 > export DEVTORCH_MCP_DOMAIN=devtorch-mcp.YOUR_ACCOUNT.workers.dev
 > /path/to/DevTorch/Implementation/deploy/cloudflare-worker/install-client-config.sh \
 >   <ide> <org> <repo> <api-key>

{agentorch-2.0.2 → agentorch-2.1.0}/agentorch.egg-info/SOURCES.txt

@@ -135,8 +135,10 @@ devtorch_core/reasoning_plus/prompt.py
 devtorch_core/reasoning_plus/learning/__init__.py
 devtorch_core/reasoning_plus/learning/api.py
 devtorch_core/reasoning_plus/learning/composer.py
+devtorch_core/reasoning_plus/learning/embeddings.py
 devtorch_core/reasoning_plus/learning/extractor.py
 devtorch_core/reasoning_plus/learning/models.py
+devtorch_core/reasoning_plus/learning/provenance.py
 devtorch_core/reasoning_plus/learning/recorder.py
 devtorch_core/reasoning_plus/learning/relevance.py
 devtorch_core/reasoning_plus/learning/state.py

{agentorch-2.0.2 → agentorch-2.1.0}/agentorch.egg-info/requires.txt

@@ -1,6 +1,6 @@
 
 [all]
-agentorch[cloud,mcp,proxy,wrapper]
+agentorch[cloud,embeddings,mcp,proxy,wrapper]
 
 [broadcast]
 watchdog>=3.0.0
@@ -18,6 +18,9 @@ azure-storage-blob>=12.19.0
 pytest>=8.0.0
 boto3>=1.34.0
 
+[embeddings]
+sentence-transformers>=3.0.0
+
 [mcp]
 mcp>=1.0.0
 

{agentorch-2.0.2 → agentorch-2.1.0}/devtorch_cli/main.py

@@ -1892,6 +1892,10 @@ def build_parser() -> argparse.ArgumentParser:
     rp_config.add_argument("--learning-enabled", type=lambda x: x.lower() in ("1", "true", "yes", "on"), default=None)
     rp_config.add_argument("--learning-top-n", type=int, default=None)
     rp_config.add_argument("--learning-max-lines", type=int, default=None)
+    rp_config.add_argument("--learning-embedding-backend", type=str, default=None)
+    rp_config.add_argument("--learning-embedding-model", type=str, default=None)
+    rp_config.add_argument("--learning-extraction-strategy", type=str, default=None)
+    rp_config.add_argument("--learning-relevance-cache-ttl", type=int, default=None)
     rp_config.set_defaults(func=cmd_reasoning_plus_config)
     rp_reset = rp_sub.add_parser("reset", help="Reset project Reasoning Plus settings to default")
     rp_reset.add_argument("--path", default=".", help="Path to project root (default: .)")
@@ -2164,6 +2168,10 @@ def cmd_reasoning_plus_status(args: argparse.Namespace) -> int:
     print(f"  learning_enabled: {config.learning_enabled}")
     print(f"  learning_top_n: {config.learning_top_n}")
     print(f"  learning_max_lines: {config.learning_max_lines}")
+    print(f"  learning_embedding_backend: {config.learning_embedding_backend}")
+    print(f"  learning_embedding_model: {config.learning_embedding_model}")
+    print(f"  learning_extraction_strategy: {config.learning_extraction_strategy}")
+    print(f"  learning_relevance_cache_ttl: {config.learning_relevance_cache_ttl}")
     return 0
 
 
@@ -2192,6 +2200,14 @@ def cmd_reasoning_plus_config(args: argparse.Namespace) -> int:
         overrides["learning_top_n"] = args.learning_top_n
     if args.learning_max_lines is not None:
         overrides["learning_max_lines"] = args.learning_max_lines
+    if args.learning_embedding_backend is not None:
+        overrides["learning_embedding_backend"] = args.learning_embedding_backend
+    if args.learning_embedding_model is not None:
+        overrides["learning_embedding_model"] = args.learning_embedding_model
+    if args.learning_extraction_strategy is not None:
+        overrides["learning_extraction_strategy"] = args.learning_extraction_strategy
+    if args.learning_relevance_cache_ttl is not None:
+        overrides["learning_relevance_cache_ttl"] = args.learning_relevance_cache_ttl
     config = config.with_values(**overrides)
     store.save_project(config)
     print("Reasoning Plus configuration updated.")

{agentorch-2.0.2 → agentorch-2.1.0}/devtorch_core/mcp/server.py

@@ -178,6 +178,10 @@ TOOLS: List[Dict[str, Any]] = [
                 "learning_enabled": {"type": "boolean", "description": "Enable or disable Reasoning Plus Learning"},
                 "learning_top_n": {"type": "integer", "description": "Number of learnings to inject"},
                 "learning_max_lines": {"type": "integer", "description": "Max lines for the injected learnings block"},
+                "learning_embedding_backend": {"type": "string", "description": "Embedding backend for relevance: keyword, sentence_transformers, openai"},
+                "learning_embedding_model": {"type": "string", "description": "Model name for sentence_transformers or openai backend"},
+                "learning_extraction_strategy": {"type": "string", "description": "Learning extraction strategy: rule or llm"},
+                "learning_relevance_cache_ttl": {"type": "integer", "description": "Relevance cache TTL in seconds"},
             },
         },
     },

{agentorch-2.0.2 → agentorch-2.1.0}/devtorch_core/reasoning_plus/augmenter.py

@@ -166,17 +166,76 @@ class ReasoningPlusAugmenter:
 
         Callers should append the returned block to the user prompt.
         """
+        block, _ = self.get_learning_context_with_provenance(context, session_id=session_id)
+        return block
+
+    def get_learning_context_with_provenance(
+        self,
+        context: str,
+        session_id: str = "",
+        call_id: str = "",
+        model_name: str = "",
+    ) -> tuple[str, list[str]]:
+        """
+        Return a compact <learnings> block plus the IDs of injected learnings.
+
+        If *call_id* is provided, a provenance record is saved so outcomes can be
+        attributed back to the injected learnings.
+        """
         if not self._learning:
-            return ""
+            return "", []
         try:
             learnings = self._learning.get_relevant_learnings(
                 context,
                 top_n=self._config.learning_top_n,
             )
-            return self._learning.compose_user_prompt("", learnings).strip()
+            block = self._learning.compose_user_prompt("", learnings).strip()
+            learning_ids = [l.id for l in learnings]
+            if call_id and learning_ids:
+                self._learning.record_injection(
+                    learning_ids,
+                    call_id=call_id,
+                    session_id=session_id,
+                    prompt_text=context,
+                    model_name=model_name,
+                )
+            return block, learning_ids
         except Exception as exc:
             logger.warning("devtorch: learning context retrieval failed — %s", exc)
-            return ""
+            return "", []
+
+    def build_system_prompt_with_learning_context(
+        self,
+        base_prompt: str,
+        prompt_text_for_context: str | None = None,
+        *,
+        task_prompt: str = "",
+        call_id: str = "",
+        session_id: str = "",
+        model_name: str = "",
+    ) -> tuple[str, list[str]]:
+        """
+        Build a Reasoning Plus system prompt and append relevant learnings.
+
+        Returns the augmented prompt and the list of learning IDs that were
+        injected (if any). A provenance record is written when *call_id* is set.
+        """
+        system_prompt = self.build_system_prompt(
+            base_prompt,
+            prompt_text_for_context=prompt_text_for_context,
+            task_prompt=task_prompt,
+        )
+        learning_ids: list[str] = []
+        if prompt_text_for_context:
+            block, learning_ids = self.get_learning_context_with_provenance(
+                prompt_text_for_context,
+                session_id=session_id,
+                call_id=call_id,
+                model_name=model_name,
+            )
+            if block:
+                system_prompt = f"{system_prompt}\n\n[DevTorch learnings]\n{block}"
+        return system_prompt, learning_ids
 
     def record_learning_call(
         self,

{agentorch-2.0.2 → agentorch-2.1.0}/devtorch_core/reasoning_plus/config.py

@@ -46,6 +46,10 @@ class ReasoningPlusConfig:
     learning_enabled: bool = True
     learning_top_n: int = 3
     learning_max_lines: int = 100
+    learning_embedding_backend: str = "keyword"
+    learning_embedding_model: str | None = None
+    learning_extraction_strategy: str = "rule"
+    learning_relevance_cache_ttl: int = 60
 
     def with_values(self, **kwargs: Any) -> "ReasoningPlusConfig":
         """Return a new config with the provided field overrides."""
@@ -69,6 +73,10 @@ class ReasoningPlusConfig:
             "learning_enabled": self.learning_enabled,
             "learning_top_n": self.learning_top_n,
             "learning_max_lines": self.learning_max_lines,
+            "learning_embedding_backend": self.learning_embedding_backend,
+            "learning_embedding_model": self.learning_embedding_model,
+            "learning_extraction_strategy": self.learning_extraction_strategy,
+            "learning_relevance_cache_ttl": self.learning_relevance_cache_ttl,
         }
 
     @classmethod
@@ -85,6 +93,10 @@ class ReasoningPlusConfig:
             learning_enabled=_bool(data.get("learning_enabled", True)),
             learning_top_n=_int(data.get("learning_top_n", 3), 3),
             learning_max_lines=_int(data.get("learning_max_lines", 100), 100),
+            learning_embedding_backend=_str(data.get("learning_embedding_backend", "keyword"), "keyword"),
+            learning_embedding_model=_optional_str(data.get("learning_embedding_model")),
+            learning_extraction_strategy=_str(data.get("learning_extraction_strategy", "rule"), "rule"),
+            learning_relevance_cache_ttl=_int(data.get("learning_relevance_cache_ttl", 60), 60),
         )
 
 
@@ -193,6 +205,21 @@ def _int(value: Any, default: int) -> int:
         return default
 
 
+def _str(value: Any, default: str) -> str:
+    if isinstance(value, str):
+        return value.strip()
+    return default
+
+
+def _optional_str(value: Any) -> str | None:
+    if value is None:
+        return None
+    if isinstance(value, str):
+        value = value.strip()
+        return value if value else None
+    return None
+
+
 def _read_json(path: Path) -> dict[str, Any] | None:
     if not path.exists():
         return None