@toolbeltai/skills 1.0.8 → 1.0.10

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.
Files changed (2) hide show
  1. package/package.json +1 -1
  2. package/toolbelt/SKILL.md +31 -19
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@toolbeltai/skills",
3
- "version": "1.0.8",
3
+ "version": "1.0.10",
4
4
  "description": "Official Toolbelt skill — one /toolbelt slash command that gives any MCP-capable agent a shared brain for your data. Detects or provisions a free Toolbelt account, configures the MCP connection, and hands off to Toolbelt's MCP tools (vector, knowledge graph, SQL, geospatial, streaming, timeline).",
5
5
  "license": "Apache-2.0",
6
6
  "homepage": "https://toolbelt.ai",
package/toolbelt/SKILL.md CHANGED
@@ -1,31 +1,33 @@
1
1
  ---
2
2
  name: toolbelt
3
3
  description: >
4
- Toolbelt is the agentic data layer. Connect Claude, Cursor, ChatGPT,
5
- and other AI agents to your structured and unstructured data through a
6
- single MCP server — vector, knowledge graph, SQL, geospatial,
7
- streaming. Built by Kinetica. Toolbelt is a collaborative substrate
8
- over your data: discover documents, structured data, events, entities,
9
- and relationships across agents and sessions. Better answers. Fewer
10
- tokens. Curated context, not raw access.
11
-
12
- Use this skill at the start of any task where an agent needs persistent
13
- memory across sessions, natural-language access to structured or
14
- unstructured data, or shared state with other agents working on the
15
- same data. The skill handles first-time setup: provisions a free
16
- Toolbelt account if none exists, configures the MCP connection in the
17
- agent's client, optionally upgrades the account by email, and hands off
18
- to Toolbelt's MCP tools for the actual work.
19
-
20
- NOT for one-off lookups that don't benefit from persistent memory or
21
- structured data — use the agent's native tools for those.
4
+ Toolbelt is a collaborative substrate over your data. Upload any
5
+ document — entities and relationships extracted automatically,
6
+ queryable immediately. Ask questions that span structured tables,
7
+ documents, and relationships in a single call. No stitching databases
8
+ together. Toolbelt orchestrates semantic, structured, and hybrid
9
+ retrieval through one MCP server — vector, knowledge graph, SQL,
10
+ geospatial, streaming. Share the URL and any agent can query the same
11
+ workspace — like a shared Google Doc for your data. Built by Kinetica.
12
+
13
+ Use this skill at the start of any task where an agent needs to ingest
14
+ documents and have entities/relationships auto-extracted, query
15
+ structured + unstructured data together in natural language, or share
16
+ findings with other agents across sessions. The skill handles
17
+ first-time setup: provisions a free Toolbelt account if none exists,
18
+ configures the MCP connection in the agent's client, and hands off to
19
+ Toolbelt's MCP tools for the actual work.
20
+
21
+ NOT for one-off lookups that don't benefit from automatic extraction,
22
+ hybrid retrieval, or shared state — use the agent's native tools for
23
+ those.
22
24
  license: Apache-2.0
23
25
  compatibility: >
24
26
  Requires an MCP-compatible AI agent (Claude Code, Claude Desktop,
25
27
  OpenClaw, Cursor, Windsurf, Gemini CLI, Codex CLI, or any client that
26
28
  supports MCP server connections). No Toolbelt account required — this
27
29
  skill provisions one on first use.
28
- version: "1.0.0"
30
+ version: "1.0.1"
29
31
  metadata:
30
32
  author: toolbeltai
31
33
  homepage: "https://toolbelt.ai"
@@ -39,6 +41,16 @@ Toolbelt is a collaborative substrate over your data. Discover documents,
39
41
  structured data, events, entities, and relationships across agents and
40
42
  sessions. Better answers. Fewer tokens. Curated context, not raw access.
41
43
 
44
+ Three things make it different:
45
+
46
+ - **Knowledge extraction.** Upload any document — entities and
47
+ relationships extracted automatically, queryable immediately.
48
+ - **Hybrid retrieval.** Ask questions that span structured tables,
49
+ documents, and relationships in a single call. No stitching databases
50
+ together. Orchestrates semantic, structured, and hybrid retrieval.
51
+ - **Shared workspaces.** Share the URL and any agent can query the same
52
+ workspace — like a shared Google Doc for your data.
53
+
42
54
  ## Two surfaces — keep them straight
43
55
 
44
56
  Toolbelt has exactly two surfaces. Knowing which is which is the most