You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+26-24Lines changed: 26 additions & 24 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -43,7 +43,7 @@ pip install openkb
43
43
```
44
44
45
45
<details>
46
-
<summary><i>Other install options</i></summary>
46
+
<summary><i>Other install options:</i></summary>
47
47
48
48
-**Latest from GitHub:**
49
49
@@ -89,7 +89,7 @@ openkb skill new my-expert "Reason like an expert on <your-topic>"
89
89
90
90
OpenKB comes with [multi-LLM support](https://docs.litellm.ai/docs/providers) (e.g., OpenAI, Claude, Gemini) via [LiteLLM](https://github.com/BerriAI/litellm) (pinned to a [safe version](https://docs.litellm.ai/blog/security-update-march-2026)).
91
91
92
-
Set your model during `openkb init`, or in [`.openkb/config.yaml`](#configuration), using `provider/model` LiteLLM format (like `anthropic/claude-sonnet-4-6`). OpenAI models can omit the prefix (like `gpt-5.4`).
92
+
Set your model during `openkb init`, or in [`.openkb/config.yaml`](#configuration), using the `provider/model` LiteLLM format (like `anthropic/claude-sonnet-4-6`). OpenAI models can omit the prefix (like `gpt-5.4`).
93
93
94
94
Create a `.env` file with your LLM API key:
95
95
@@ -162,14 +162,14 @@ OpenKB commands fall into two layers: the **wiki foundation** (compile + manage
162
162
| Command | Description |
163
163
|---|---|
164
164
|`openkb init`| Initialize a new knowledge base (interactive) |
165
-
| <code>openkb add <file_or_dir_or_URL></code> | Add files, directories, or URLs and compile to wiki. URLs auto-detect content type (PDF or HTML) |
165
+
| <code>openkb add <file_or_dir_or_URL></code> | Add files, directories, or URLs and compile to wiki. URL content type (PDF or HTML) is auto-detected|
166
166
|`openkb list`| List indexed documents and concepts |
167
167
|`openkb status`| Show knowledge base stats |
168
168
|`openkb watch`| Watch `raw/` and auto-compile new files |
169
169
|`openkb lint`| Run structural + knowledge health checks |
170
170
171
171
<details>
172
-
<summary><i>More wiki commands:</i></summary>
172
+
<summary><b><i>More wiki commands:</i></b></summary>
173
173
174
174
| Command | Description |
175
175
|---|---|
@@ -193,7 +193,7 @@ A "generator" reads from the compiled wiki and produces something usable: an ans
193
193
| <code>openkb skill new <name> "<intent>"</code> | A redistributable agent skill — see [Skill Factory](#-skill-factory--drop-in-a-book-out-comes-a-digital-expert) below |
@@ -203,9 +203,9 @@ A "generator" reads from the compiled wiki and produces something usable: an ans
203
203
204
204
</details>
205
205
206
-
### Query & Chat — ask the wiki
206
+
### 💬 Query & Chat — *Ask the wiki*
207
207
208
-
`openkb query "..."` answers a single question. `openkb chat` is interactive; each turn carries history, so you can dig into a topic without re-typing context. Both use the same underlying wiki and the same retrieval primitives (PageIndex for long docs, direct concept reads for short).
208
+
`openkb query "..."` answers a single question. `openkb chat` is interactive; each turn carries history, so you can dig into a topic without re-typing context. Both use the same underlying wiki and retrieval primitives.
209
209
210
210
```bash
211
211
openkb query "What does the literature say about attention scaling?"
@@ -217,10 +217,10 @@ openkb chat --list # list all sessions
217
217
openkb chat --delete <id># delete a session
218
218
```
219
219
220
-
Inside a chat, type `/` to access slash commands (Tab to complete):
220
+
Inside a chat, type `/` to access slash commands (Tab to complete).
221
221
222
222
<details>
223
-
<summary><i>Slash commands</i></summary>
223
+
<summary><i>Slash commands:</i></summary>
224
224
225
225
-`/help` — list available commands
226
226
-`/status` — show knowledge base status
@@ -234,7 +234,7 @@ Inside a chat, type `/` to access slash commands (Tab to complete):
234
234
235
235
</details>
236
236
237
-
### 🛠 Skill Factory — *Drop in a book. Out comes a digital expert.*
237
+
### 🛠 Skill Factory — *Drop in a book; out comes a digital expert.*
238
238
239
239
The newest generator. `openkb skill new` distills an [agent skill](https://docs.claude.com/en/docs/build-with-claude/skills) from any subset of your wiki, a portable folder that major agents (Claude Code, Codex, etc.) install and load natively. Drop in a book's worth of papers; out comes a specialist that other agents can call on.
240
240
@@ -244,7 +244,7 @@ openkb skill new karpathy-thinking \
244
244
```
245
245
246
246
<details>
247
-
<summary><i>Output</i></summary>
247
+
<summary><i>Output:</i></summary>
248
248
249
249
```
250
250
<kb>/output/skills/karpathy-thinking/
@@ -260,7 +260,7 @@ openkb skill new karpathy-thinking \
Settings are initialized by `openkb init`, and stored in `.openkb/config.yaml`:
320
+
OpenKB settings are initialized by `openkb init` and are stored in `.openkb/config.yaml`:
321
321
322
322
```yaml
323
323
model: gpt-5.4 # LLM model (any LiteLLM-supported provider)
@@ -334,7 +334,7 @@ Model names use `provider/model` LiteLLM [format](https://docs.litellm.ai/docs/p
334
334
| Gemini | `gemini/gemini-3.1-pro-preview` |
335
335
336
336
<details>
337
-
<summary><i>Advanced options</i></summary>
337
+
<summary><i>Advanced options:</i></summary>
338
338
339
339
`entity_types` (optional): a YAML list overriding the entity-type vocabulary used for entity pages; omit it to use the default `person`, `organization`, `place`, `product`, `work`, `event`, `other`.
340
340
@@ -350,13 +350,13 @@ Subscription-based providers that authenticate via OAuth device flow (e.g. `chat
350
350
351
351
</details>
352
352
353
-
### PageIndex
353
+
### PageIndex Setup
354
354
355
-
Long-document retrieval is a [known challenge](https://x.com/karpathy/status/2039823314982744522) for LLMs. [PageIndex](https://github.com/VectifyAI/PageIndex) solves this with vectorless, reasoning-based retrieval, building a hierarchical tree index that lets LLMs reason over the index for context-aware retrieval.
355
+
Long-document retrieval is a [known challenge](https://x.com/karpathy/status/2039823314982744522) for LLMs. [PageIndex](https://github.com/VectifyAI/PageIndex) solves this with vectorless, reasoning-based retrieval, by building a hierarchical tree index that lets LLMs reason over the index for context-aware retrieval.
356
356
357
357
PageIndex runs locally by default using the [open-source version](https://github.com/VectifyAI/PageIndex), with no external dependencies required.
358
358
359
-
#### Cloud Support(Optional)
359
+
***Cloud Support*** *(Optional)*:
360
360
361
361
For large or complex PDFs, [PageIndex Cloud](https://docs.pageindex.ai/) can be used to access additional capabilities, including:
362
362
@@ -392,7 +392,7 @@ OpenKB's wiki is a directory of Markdown files with `[[wikilinks]]`. Obsidian re
392
392
OpenKB ships a `SKILL.md` so any agent CLI can read your compiled wiki. No extra runtime, no MCP setup, just install the skill once.
393
393
394
394
<details>
395
-
<summary><b>Claude Code</b></summary>
395
+
<summary><i>Claude Code:</i></summary>
396
396
397
397
```
398
398
/plugin marketplace add VectifyAI/OpenKB
@@ -402,7 +402,9 @@ OpenKB ships a `SKILL.md` so any agent CLI can read your compiled wiki. No extra
0 commit comments