@runcontext/cli 0.4.0 → 0.4.1

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 CHANGED
@@ -1,102 +1,66 @@
1
1
  # @runcontext/cli
2
2
 
3
- CLI for [ContextKit](https://github.com/erickittelson/ContextKit) — AI-ready metadata governance over OSI.
3
+ **ContextKit — tell your AI agent to build your semantic layer.**
4
4
 
5
- **Your data already has a schema. ContextKit gives it meaning.**
5
+ Tell your AI agent: *"Install @runcontext/cli and build a semantic layer for my database."*
6
+
7
+ The agent introspects your database, scaffolds metadata, and goes back and forth with you — asking about metrics, ownership, and business rules — while it builds the semantic layer using CLI commands. When it reaches Gold tier, it exports an **AI Blueprint** and serves the metadata to other AI agents via MCP.
6
8
 
7
9
  ## Installation
8
10
 
9
11
  ```bash
10
- # Global install
11
- npm install -g @runcontext/cli
12
-
13
- # Or per-project
14
- npm install -D @runcontext/cli
12
+ npm install @runcontext/cli
15
13
  ```
16
14
 
17
15
  ## Quick Start
18
16
 
19
- ```bash
20
- # Point at any database — interactive wizard does the rest
21
- context setup
22
-
23
- # Or step by step
24
- context introspect --db duckdb://warehouse.duckdb
25
- context enrich --target silver --apply
26
- context tier
27
- ```
17
+ In Claude Code, Cursor, Windsurf, or any agentic coding platform:
28
18
 
29
- ## Commands
19
+ > *"Install @runcontext/cli and build a semantic layer for my database."*
30
20
 
31
- ### Core Workflow
21
+ Or run it yourself:
32
22
 
33
23
  ```bash
34
- context setup # Interactive wizard — database to metadata in one flow
35
- context introspect # Scan a database → scaffold Bronze-level OSI metadata
36
- context enrich --target silver # Auto-enrich metadata toward a target tier
37
- context lint # Run 37 lint rules against context files
38
- context fix --write # Auto-fix lint issues where possible
39
- context build # Compile context files → emit manifest JSON
40
- context tier [model] # Show Bronze/Silver/Gold scorecard
24
+ context setup # Interactive wizard — database to metadata in one flow
25
+ context tier # Check Bronze/Silver/Gold score
26
+ context blueprint # Export AI Blueprint (portable Gold-tier spec)
27
+ context serve --stdio # Serve to AI agents via MCP
41
28
  ```
42
29
 
43
- ### Exploration
30
+ ## Commands
44
31
 
45
32
  ```bash
46
- context explain <name> # Look up any model, term, or owner
47
- context rules # List all lint rules with tier, severity, fixability
48
- context validate-osi <file> # Validate a single OSI file against the schema
49
- context verify # Check metadata accuracy against a live database
50
- ```
51
-
52
- ### Serving
33
+ # Build the semantic layer
34
+ context setup # Interactive wizard full pipeline in one flow
35
+ context new <name> # Scaffold a new data product
36
+ context introspect # Scan a database -> scaffold Bronze metadata
37
+ context enrich --target silver # Auto-enrich toward a target tier
38
+ context lint # Run 40 lint rules
39
+ context fix --write # Auto-fix lint issues
40
+ context build # Compile -> emit manifest JSON
41
+ context tier [model] # Show Bronze/Silver/Gold scorecard
53
42
 
54
- ```bash
55
- context serve --stdio # Start MCP server over stdio (for Claude, Cursor, etc.)
56
- context serve --http --port 3000 # Start MCP server over HTTP
57
- context site # Generate a static documentation site
58
- context dev # Watch mode re-lint on file changes
59
- context init # Scaffold a new project structure
43
+ # Explore and verify
44
+ context explain <name> # Look up any model, term, or owner
45
+ context rules # List all lint rules
46
+ context validate-osi <file> # Validate against OSI spec
47
+ context verify # Check accuracy against a live database
48
+
49
+ # Export and serve
50
+ context blueprint [model] # Export AI Blueprints (portable OSI YAML)
51
+ context serve --stdio # MCP server over stdio
52
+ context serve --http --port 3000 # MCP server over HTTP
53
+ context site # Static documentation site
54
+ context dev --studio # Visual editor in the browser
55
+ context init # Scaffold a new project
60
56
  ```
61
57
 
62
- ## The Tier System
63
-
64
- | Tier | What it means | How to get there |
65
- |---|---|---|
66
- | **Bronze** | Discoverable — described, owned, classified | `context introspect` |
67
- | **Silver** | Trusted — lineage, glossary, sample values, trust status | `context enrich --target silver` |
68
- | **Gold** | AI-Ready — semantic roles, golden queries, guardrails, business rules | Human curation + `context enrich --target gold` |
69
-
70
58
  ## Database Support
71
59
 
72
- | Adapter | Connection |
73
- |---|---|
74
- | DuckDB | `--db duckdb://path.duckdb` |
75
- | PostgreSQL | `--db postgres://user:pass@host:5432/db` |
76
- | MySQL | `--db mysql://user:pass@host:3306/db` |
77
- | SQL Server | `--db mssql://user:pass@host:1433/db` |
78
- | SQLite | `--db path/to/file.sqlite` |
79
- | Snowflake | `--db snowflake://account/database/schema` |
80
- | BigQuery | `--db bigquery://project/dataset` |
81
- | ClickHouse | `--db clickhouse://host:8123` |
82
- | Databricks | Config file only |
83
-
84
- Each adapter requires its own driver as an optional peer dependency. See [Database Support docs](https://contextkit.dev/reference/databases/) for installation details.
60
+ DuckDB, PostgreSQL, MySQL, SQL Server, SQLite, Snowflake, BigQuery, ClickHouse, Databricks.
85
61
 
86
62
  ## MCP Server
87
63
 
88
- Expose your metadata to AI agents:
89
-
90
- ```bash
91
- # For Claude Code / Cursor
92
- context serve --stdio
93
-
94
- # For multi-agent setups
95
- context serve --http --port 3000
96
- ```
97
-
98
- Add to `.claude/mcp.json`:
99
-
100
64
  ```json
101
65
  {
102
66
  "mcpServers": {
@@ -108,9 +72,9 @@ Add to `.claude/mcp.json`:
108
72
  }
109
73
  ```
110
74
 
111
- ## Full Documentation
75
+ ## Documentation
112
76
 
113
- See the [ContextKit repository](https://github.com/erickittelson/ContextKit) for complete docs, file structure, tier requirements, and examples.
77
+ [contextkit.dev](https://contextkit.dev) | [GitHub](https://github.com/erickittelson/ContextKit)
114
78
 
115
79
  ## License
116
80
 
package/dist/index.js CHANGED
@@ -141,7 +141,7 @@ function formatSarif(diagnostics) {
141
141
  tool: {
142
142
  driver: {
143
143
  name: "ContextKit",
144
- version: "0.4.0",
144
+ version: "0.4.1",
145
145
  informationUri: "https://github.com/erickittelson/ContextKit",
146
146
  rules: Array.from(ruleMap.values())
147
147
  }
@@ -579,9 +579,9 @@ function parseDbUrl(db) {
579
579
  `Cannot determine adapter from "${db}". Use a URL prefix (duckdb://, postgres://, mysql://, mssql://, clickhouse://, snowflake://, bigquery://) or a recognized file extension (.duckdb, .db, .sqlite, .sqlite3).`
580
580
  );
581
581
  }
582
- var introspectCommand = new Command5("introspect").description("Introspect a database and scaffold Bronze-level OSI metadata").option(
582
+ var introspectCommand = new Command5("introspect").description("Introspect a database and scaffold Bronze-level OSI metadata. Supports: duckdb://, postgres://, mysql://, mssql://, snowflake://, bigquery://, clickhouse://, .sqlite, .duckdb files, and Databricks (via config).").option(
583
583
  "--db <url>",
584
- "Database URL (e.g., duckdb://path.duckdb or postgres://...)"
584
+ "Database URL (duckdb://path.duckdb, postgres://user:pass@host/db, mysql://..., mssql://..., snowflake://account/db/schema, bigquery://project/dataset, clickhouse://host, or file.sqlite)"
585
585
  ).option(
586
586
  "--source <name>",
587
587
  "Use a named data_source from contextkit.config.yaml"
@@ -3638,7 +3638,7 @@ async function runAgentInstructionsStep(ctx) {
3638
3638
  }
3639
3639
 
3640
3640
  // src/commands/setup.ts
3641
- var setupCommand = new Command15("setup").description("Interactive wizard to scaffold and enrich metadata from a database").action(async () => {
3641
+ var setupCommand = new Command15("setup").description("Interactive wizard \u2014 detects databases, introspects schema, scaffolds metadata, enriches to Silver, generates agent instructions. Supports DuckDB, PostgreSQL, MySQL, SQL Server, SQLite, Snowflake, BigQuery, ClickHouse, and Databricks.").action(async () => {
3642
3642
  p10.intro(chalk16.bgCyan(chalk16.black(" ContextKit Setup ")));
3643
3643
  const ctx = await runConnectStep();
3644
3644
  if (!ctx) return;
@@ -4131,7 +4131,7 @@ var newCommand = new Command17("new").description("Scaffold a new data product i
4131
4131
 
4132
4132
  // src/index.ts
4133
4133
  var program = new Command18();
4134
- program.name("context").description("ContextKit \u2014 AI-ready metadata governance over OSI").version("0.4.0");
4134
+ program.name("context").description("ContextKit \u2014 AI-ready metadata governance over OSI").version("0.4.1");
4135
4135
  program.addCommand(lintCommand);
4136
4136
  program.addCommand(buildCommand);
4137
4137
  program.addCommand(tierCommand);