@hermenics/deepseek-code 0.2.17 → 0.3.1
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 +7 -4
- package/dist/cli.mjs +473 -503
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -30,7 +30,8 @@
|
|
|
30
30
|
- **Agentic coding** — reads/writes files, runs shell commands, searches code, manages git
|
|
31
31
|
- **Multi-provider** — DeepSeek API, Amazon Bedrock, Google Vertex AI, or any local model (Ollama, LM Studio)
|
|
32
32
|
- **Full TUI** — alternate-screen interface with streamed thinking, rich markdown, and vim mode
|
|
33
|
-
- **
|
|
33
|
+
- **Unified agents** — scoped primary agents and sub-agents with editable base prompts
|
|
34
|
+
- **Settings center** — fullscreen, searchable User/Project/Local configuration with visible origins
|
|
34
35
|
- **MCP support** — connect to any Model Context Protocol server for extended tooling
|
|
35
36
|
- **Extensible** — slash commands, custom tools, memory, sessions, themes
|
|
36
37
|
|
|
@@ -63,7 +64,7 @@ cat src/index.tsx | deepseek --pipe --json "summarize"
|
|
|
63
64
|
| **Google Vertex AI** | GCP service account JSON key | `GCP_PROJECT`, `GCP_LOCATION`, `GCP_CREDENTIALS` |
|
|
64
65
|
| **Local (Ollama / LM Studio)** | No auth — point to your local endpoint | `LOCAL_BASE_URL`, `LOCAL_MODEL` |
|
|
65
66
|
|
|
66
|
-
|
|
67
|
+
Secrets are saved only to `~/.deepseek/config.json`. Non-secret preferences use `settings.json` with `User < Project < Local` precedence; legacy values remain readable for compatibility. See [docs/settings.md](docs/settings.md).
|
|
67
68
|
|
|
68
69
|
## Models
|
|
69
70
|
|
|
@@ -89,6 +90,7 @@ Each provider also exposes provider-specific models (Bedrock, Vertex, local).
|
|
|
89
90
|
| `/theme` | Change color theme |
|
|
90
91
|
| `/tools` | List available tools |
|
|
91
92
|
| `/permissions` | Explain mode, allow/deny rules, risk checks, and session approvals |
|
|
93
|
+
| `/config`, `/settings` | Open the fullscreen settings center |
|
|
92
94
|
| `/help` | Show all commands |
|
|
93
95
|
|
|
94
96
|
## Built-in tools
|
|
@@ -99,9 +101,10 @@ The agent has access to these tools out of the box:
|
|
|
99
101
|
|
|
100
102
|
## TUI behavior
|
|
101
103
|
|
|
102
|
-
-
|
|
104
|
+
- Uses the terminal **alternate screen** by default; it can be disabled for the next session in `/settings`
|
|
105
|
+
- Settings adapt from three panes to a sequential category → list → detail flow on narrow terminals
|
|
103
106
|
- Thinking output is streamed as full multiline blocks and persisted after each response
|
|
104
|
-
-
|
|
107
|
+
- Build, Plan, Review and Auto are real interaction modes. Review is read-only; Plan can only write its designated plan.
|
|
105
108
|
|
|
106
109
|
---
|
|
107
110
|
|