claude-smart 0.2.28 → 0.2.30
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +23 -16
- package/bin/claude-smart.js +333 -73
- package/package.json +1 -1
- package/plugin/.claude-plugin/plugin.json +1 -1
- package/plugin/.codex-plugin/plugin.json +1 -1
- package/plugin/README.md +4 -0
- package/plugin/hooks/codex-hooks.json +8 -3
- package/plugin/hooks/hooks.json +13 -3
- package/plugin/pyproject.toml +1 -1
- package/plugin/scripts/_lib.sh +38 -0
- package/plugin/scripts/backend-log-runner.sh +33 -0
- package/plugin/scripts/backend-service.sh +24 -12
- package/plugin/scripts/cli.sh +27 -3
- package/plugin/scripts/codex-claude-compat +9 -0
- package/plugin/scripts/codex-claude-compat.cmd +4 -0
- package/plugin/scripts/codex-claude-compat.js +162 -0
- package/plugin/scripts/codex-hook.js +102 -11
- package/plugin/scripts/dashboard-service.sh +13 -4
- package/plugin/scripts/hook_entry.sh +29 -3
- package/plugin/scripts/smart-install.sh +176 -50
- package/plugin/src/claude_smart/cli.py +101 -2
- package/plugin/src/claude_smart/context_inject.py +2 -4
- package/plugin/src/claude_smart/cs_cite.py +2 -90
- package/plugin/src/claude_smart/events/stop.py +16 -42
- package/plugin/src/claude_smart/internal_call.py +23 -0
- package/plugin/src/claude_smart/state.py +3 -3
- package/plugin/uv.lock +73 -76
- package/plugin/bin/cs-cite +0 -77
- package/plugin/scripts/codex-claude-compat.py +0 -144
package/README.md
CHANGED
|
@@ -6,14 +6,14 @@
|
|
|
6
6
|
claude-smart
|
|
7
7
|
</h1>
|
|
8
8
|
|
|
9
|
-
<h4 align="center">A
|
|
9
|
+
<h4 align="center">A self-improvement plugin for <a href="https://claude.com/claude-code" target="_blank">Claude Code</a> and Codex that turns interactions into durable skills they follow in future sessions.</h4>
|
|
10
10
|
|
|
11
11
|
<p align="center">
|
|
12
12
|
<a href="LICENSE">
|
|
13
13
|
<img src="https://img.shields.io/badge/License-Apache%202.0-blue.svg" alt="License">
|
|
14
14
|
</a>
|
|
15
15
|
<a href="plugin/pyproject.toml">
|
|
16
|
-
<img src="https://img.shields.io/badge/version-0.2.
|
|
16
|
+
<img src="https://img.shields.io/badge/version-0.2.30-green.svg" alt="Version">
|
|
17
17
|
</a>
|
|
18
18
|
<a href="plugin/pyproject.toml">
|
|
19
19
|
<img src="https://img.shields.io/badge/python-%3E%3D3.12-brightgreen.svg" alt="Python">
|
|
@@ -24,9 +24,6 @@
|
|
|
24
24
|
<a href="#quick-start">
|
|
25
25
|
<img src="https://img.shields.io/badge/hosts-Claude%20Code%20%2B%20Codex-purple.svg" alt="Hosts">
|
|
26
26
|
</a>
|
|
27
|
-
<a href="https://discord.gg/Jbft3jPn">
|
|
28
|
-
<img src="https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white" alt="Discord">
|
|
29
|
-
</a>
|
|
30
27
|
</p>
|
|
31
28
|
|
|
32
29
|
<p align="center">
|
|
@@ -40,22 +37,22 @@
|
|
|
40
37
|
</p>
|
|
41
38
|
|
|
42
39
|
<p align="center">
|
|
43
|
-
|
|
40
|
+
claude-smart learns from your corrections <i>and</i> from what's already working — so the assistant stops repeating mistakes and reuses proven paths. Project-specific skills capture repo-local rules, and shared skills roll up durable patterns for reuse across projects.
|
|
44
41
|
</p>
|
|
45
42
|
|
|
46
43
|
<p align="center">
|
|
47
|
-
<b>vs <code>claude-mem</code>:</b> ~3× better at turning your corrections into rules Claude follows, ~50% more of
|
|
44
|
+
<b>vs <code>claude-mem</code>:</b> ~3× better at turning your corrections into rules Claude follows, ~50% more of your guidance is retained. <a href="EXPERIMENT.md"><b>Read the benchmark →</b></a>
|
|
48
45
|
</p>
|
|
49
46
|
|
|
50
47
|
---
|
|
51
48
|
|
|
52
49
|
## Why Learning, Not Memory
|
|
53
50
|
|
|
54
|
-
Most memory
|
|
51
|
+
Most memory tools just record. Claude remembers what happened, but doesn't change what it does next.
|
|
55
52
|
|
|
56
53
|
`claude-smart` focuses on learning instead.
|
|
57
54
|
|
|
58
|
-
Four
|
|
55
|
+
Four things this changes:
|
|
59
56
|
|
|
60
57
|
- 💡 **Stop repeating the same mistakes:** Produces actionable skills Claude can follow next time; memory only records what happened.
|
|
61
58
|
|
|
@@ -63,12 +60,12 @@ Four ways this changes what your coding assistant can do for you:
|
|
|
63
60
|
> **Memory:** “deploy broke after prisma bump; user rolled back”<br>
|
|
64
61
|
> **Learning:** “treat major-version bumps of ORMs/DB drivers as breaking — verify with integration tests, not just unit tests”
|
|
65
62
|
|
|
66
|
-
- 🚀 **Start from the optimized path:**
|
|
63
|
+
- 🚀 **Start from the optimized path:** Records and optimizes the steps that worked so Claude can reuse them, instead of re-exploring each session.
|
|
67
64
|
> *Example:* Claude spends several iterations trying to start the local dev environment before discovering that this repo requires `pnpm dev:all` instead of the usual `npm run dev`.<br>
|
|
68
65
|
> **Memory:** “user mentioned that `npm run dev` did not work”<br>
|
|
69
66
|
> **Learning:** “for this repo, always use `pnpm dev:all` to start the full local stack — `npm run dev` only starts the frontend and causes missing service errors”
|
|
70
67
|
|
|
71
|
-
Instead of re-exploring, Claude starts from the proven path—reducing planning steps, latency, and token usage.
|
|
68
|
+
Instead of re-exploring, Claude Code starts from the proven path—reducing planning steps, latency, and token usage.
|
|
72
69
|
|
|
73
70
|
- 🌐 **Project-specific and shared skills:** Session memory disappears with the conversation. Project-specific skills preserve repo-local rules, while shared skills roll up patterns that should transfer across projects. Preferences stay scoped to the current project so per-repo preferences don't leak across projects.
|
|
74
71
|
|
|
@@ -95,10 +92,6 @@ claude plugin marketplace add ReflexioAI/claude-smart
|
|
|
95
92
|
claude plugin install claude-smart@reflexioai
|
|
96
93
|
```
|
|
97
94
|
|
|
98
|
-
The plugin Setup hook installs its own `uv`, Python 3.12 environment, and
|
|
99
|
-
private Node.js/npm runtime under `~/.claude-smart/` when they are missing, so
|
|
100
|
-
you do not need to install Python or Node globally for the plugin/dashboard.
|
|
101
|
-
|
|
102
95
|
To uninstall:
|
|
103
96
|
|
|
104
97
|
```bash
|
|
@@ -133,6 +126,19 @@ Developing the plugin itself? See [DEVELOPER.md](./DEVELOPER.md#developing-local
|
|
|
133
126
|
|
|
134
127
|
> **Not supported:** Claude Code Cowork, claude.ai/code web, or remote Codex environments without local plugin hooks — they run outside your local machine, so the local backend/dashboard and `~/.reflexio/` aren't reachable.
|
|
135
128
|
|
|
129
|
+
### Vanilla OS support
|
|
130
|
+
|
|
131
|
+
| Platform | Status | Notes |
|
|
132
|
+
| --- | --- | --- |
|
|
133
|
+
| Apple Silicon macOS 14+ | Supported | Runtime bootstrap installs private Node/npm, uv, Python 3.12 deps, and dashboard deps. |
|
|
134
|
+
| Windows x64 | Supported | Runtime bootstrap uses PowerShell for uv/Node archive extraction and patches Codex hooks to the Node wrapper. |
|
|
135
|
+
| Linux | Supported when host hooks are local | Existing Linux behavior is preserved; install coverage depends on available Python wheels. |
|
|
136
|
+
| Intel Mac, macOS 13 or older, Windows ARM | Not supported | Current local embedding/ML dependencies do not publish a complete native wheel set for these targets. |
|
|
137
|
+
|
|
138
|
+
Network access is required during first install for npm, PyPI/uv, Node.js, and
|
|
139
|
+
the first local embedding model download. The ONNX model cache lives at
|
|
140
|
+
`~/.cache/chroma/onnx_models/all-MiniLM-L6-v2/`.
|
|
141
|
+
|
|
136
142
|
---
|
|
137
143
|
|
|
138
144
|
## Key Features
|
|
@@ -150,7 +156,7 @@ Developing the plugin itself? See [DEVELOPER.md](./DEVELOPER.md#developing-local
|
|
|
150
156
|
|
|
151
157
|
## Dashboard
|
|
152
158
|
|
|
153
|
-
A web UI for browsing session histories, inspecting preferences, and editing project-specific and shared skills. The dashboard auto-starts alongside the backend, so you can open **http://localhost:3001** directly. Or run `/claude-smart:dashboard` in Claude Code to
|
|
159
|
+
A web UI for browsing session histories, inspecting preferences, and editing project-specific and shared skills. The dashboard auto-starts alongside the backend, so you can open **http://localhost:3001** directly. Or run `/claude-smart:dashboard` in Claude Code to open it in your browser. In Codex, run `bash ~/.reflexio/plugin-root/scripts/dashboard-open.sh`.
|
|
154
160
|
|
|
155
161
|
<p align="center">
|
|
156
162
|
<img src="assets/preferences_dashboard.png" alt="Preferences dashboard" width="49%">
|
|
@@ -216,6 +222,7 @@ Advanced users can tune claude-smart via environment variables — see [DEVELOPE
|
|
|
216
222
|
| `~/.codex/plugins/cache/reflexioai/claude-smart/<version>/` | Codex's cached install of the `claude-smart` plugin from the `ReflexioAI` marketplace. |
|
|
217
223
|
| `~/.reflexio/plugin-root` | Self-healed symlink to the active plugin dir (managed by `ensure-plugin-root.sh` — written on install, refreshed each `SessionStart`). Claude Code slash commands and Codex shell-command helpers resolve through it, so don't delete it; if you do, the next session will recreate it. |
|
|
218
224
|
| `~/.claude-smart/sessions/{session_id}.jsonl` | Per-session buffer. User turns, assistant turns, tool invocations, `{"published_up_to": N}` watermarks. Safe to inspect and safe to delete — everything past the latest watermark has already been written to reflexio's DB. |
|
|
225
|
+
| `~/.claude-smart/node/current/` | Private Node.js/npm runtime used by hooks and the dashboard after install. |
|
|
219
226
|
| `~/.cache/chroma/onnx_models/all-MiniLM-L6-v2/` | Cached ONNX weights (~86 MB, downloaded once). Delete to force a re-download. |
|
|
220
227
|
|
|
221
228
|
For troubleshooting, see [TROUBLESHOOTING.md](./TROUBLESHOOTING.md).
|