@nodesify/astria 1.0.0 → 1.0.2

package/skills/skill-copilot.md

@@ -1,46 +1,46 @@
----
-name: astria
-description: Turn any directory into a queryable knowledge graph. Trigger: /astria
----
-
-# astria skill (GitHub Copilot)
-
-When the user types `/astria`, run the astria knowledge graph pipeline.
-
-## Step 1 - Build or update the graph
-
-Run via terminal:
-```bash
-node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
-```
-
-- `missing` → run `astria run .`
-- `stale` → run `astria update .`
-- `fresh` → skip to Step 2
-
-## Step 2 - Read the report
-
-Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
-## Prefer MCP tools when connected
-
-If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
-
-
-## Enforcement Rules
-
-When `.astria/` exists, you MUST use astria commands BEFORE searching files:
-
-| Question | Command |
-|----------|---------|
-| Where is X? | `astria query "X"` |
-| How does X connect to Y? | `astria path "X" "Y"` |
-| What breaks if X changes? | `astria affected "X"` |
-| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
-| What does X do? | `astria explain "X"` |
-| Architecture overview? | Read `.astria/graph_report.md` |
-
-Copilot Chat with `.github/copilot-instructions.md` will reference these rules.
-
-## After editing code
-
-Run `astria update .` to keep the graph current.
+---
+name: astria
+description: Turn any directory into a queryable knowledge graph. Trigger: /astria
+---
+
+# astria skill (GitHub Copilot)
+
+When the user types `/astria`, run the astria knowledge graph pipeline.
+
+## Step 1 - Build or update the graph
+
+Run via terminal:
+```bash
+node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
+```
+
+- `missing` → run `astria run .`
+- `stale` → run `astria update .`
+- `fresh` → skip to Step 2
+
+## Step 2 - Read the report
+
+Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
+## Prefer MCP tools when connected
+
+If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
+
+
+## Enforcement Rules
+
+When `.astria/` exists, you MUST use astria commands BEFORE searching files:
+
+| Question | Command |
+|----------|---------|
+| Where is X? | `astria query "X"` |
+| How does X connect to Y? | `astria path "X" "Y"` |
+| What breaks if X changes? | `astria affected "X"` |
+| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
+| What does X do? | `astria explain "X"` |
+| Architecture overview? | Read `.astria/graph_report.md` |
+
+Copilot Chat with `.github/copilot-instructions.md` will reference these rules.
+
+## After editing code
+
+Run `astria update .` to keep the graph current.

package/skills/skill-cursor.md

@@ -1,39 +1,39 @@
----
-name: astria
-description: Turn any directory into a queryable knowledge graph. Trigger: /astria
----
-
-# astria skill (Cursor)
-
-When the user types `/astria`, run the astria knowledge graph pipeline.
-
-## Step 1 - Build or update the graph
-
-Run via terminal:
-```bash
-node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
-```
-
-- `missing` → run `astria run .`
-- `stale` → run `astria update .`
-- `fresh` → skip to Step 2
-
-## Step 2 - Read the report
-
-Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
-## Prefer MCP tools when connected
-
-If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
-
-
-## Enforcement Rules
-
-When `.astria/` exists, the `.cursor/rules/astria.mdc` rule enforces graph usage. You MUST:
-
-1. Read `.astria/graph_report.md` before searching files
-2. Use `astria query`, `path` (add `--directed` for call direction), `explain`, or `affected` for cross-module questions
-3. Only use native file tools AFTER the graph identified the exact files
-
-## After editing code
-
-Run `astria update .` to keep the graph current.
+---
+name: astria
+description: Turn any directory into a queryable knowledge graph. Trigger: /astria
+---
+
+# astria skill (Cursor)
+
+When the user types `/astria`, run the astria knowledge graph pipeline.
+
+## Step 1 - Build or update the graph
+
+Run via terminal:
+```bash
+node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
+```
+
+- `missing` → run `astria run .`
+- `stale` → run `astria update .`
+- `fresh` → skip to Step 2
+
+## Step 2 - Read the report
+
+Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
+## Prefer MCP tools when connected
+
+If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
+
+
+## Enforcement Rules
+
+When `.astria/` exists, the `.cursor/rules/astria.mdc` rule enforces graph usage. You MUST:
+
+1. Read `.astria/graph_report.md` before searching files
+2. Use `astria query`, `path` (add `--directed` for call direction), `explain`, or `affected` for cross-module questions
+3. Only use native file tools AFTER the graph identified the exact files
+
+## After editing code
+
+Run `astria update .` to keep the graph current.

package/skills/skill-gemini.md

@@ -1,46 +1,46 @@
----
-name: astria
-description: Turn any directory into a queryable knowledge graph. Trigger: /astria
----
-
-# astria skill (Gemini)
-
-When the user types `/astria`, run the astria knowledge graph pipeline.
-
-## Step 1 - Build or update the graph
-
-Run via shell:
-```bash
-node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
-```
-
-- `missing` → run `astria run .`
-- `stale` → run `astria update .`
-- `fresh` → skip to Step 2
-
-## Step 2 - Read the report
-
-Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
-## Prefer MCP tools when connected
-
-If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
-
-
-## Enforcement Rules
-
-When `.astria/` exists, you MUST use astria commands BEFORE using `read_file` or `list_directory` for architecture questions:
-
-| Question | Command |
-|----------|---------|
-| Where is X? | `astria query "X"` |
-| How does X connect to Y? | `astria path "X" "Y"` |
-| What breaks if X changes? | `astria affected "X"` |
-| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
-| What does X do? | `astria explain "X"` |
-| Architecture overview? | Read `.astria/graph_report.md` |
-
-Only use native file tools AFTER the graph identified the exact files.
-
-## After editing code
-
-Run `astria update .` to keep the graph current.
+---
+name: astria
+description: Turn any directory into a queryable knowledge graph. Trigger: /astria
+---
+
+# astria skill (Gemini)
+
+When the user types `/astria`, run the astria knowledge graph pipeline.
+
+## Step 1 - Build or update the graph
+
+Run via shell:
+```bash
+node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
+```
+
+- `missing` → run `astria run .`
+- `stale` → run `astria update .`
+- `fresh` → skip to Step 2
+
+## Step 2 - Read the report
+
+Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
+## Prefer MCP tools when connected
+
+If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
+
+
+## Enforcement Rules
+
+When `.astria/` exists, you MUST use astria commands BEFORE using `read_file` or `list_directory` for architecture questions:
+
+| Question | Command |
+|----------|---------|
+| Where is X? | `astria query "X"` |
+| How does X connect to Y? | `astria path "X" "Y"` |
+| What breaks if X changes? | `astria affected "X"` |
+| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
+| What does X do? | `astria explain "X"` |
+| Architecture overview? | Read `.astria/graph_report.md` |
+
+Only use native file tools AFTER the graph identified the exact files.
+
+## After editing code
+
+Run `astria update .` to keep the graph current.

package/skills/skill-kiro.md

@@ -1,39 +1,39 @@
----
-name: astria
-description: Turn any directory into a queryable knowledge graph. Trigger: /astria
----
-
-# astria skill (Kiro)
-
-When the user types `/astria`, run the astria knowledge graph pipeline.
-
-## Step 1 - Build or update the graph
-
-Run via terminal:
-```bash
-node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
-```
-
-- `missing` → run `astria run .`
-- `stale` → run `astria update .`
-- `fresh` → skip to Step 2
-
-## Step 2 - Read the report
-
-Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
-## Prefer MCP tools when connected
-
-If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
-
-
-## Enforcement Rules
-
-The `.kiro/steering/astria.md` steering file enforces graph usage. You MUST:
-
-1. Read `.astria/graph_report.md` before searching files
-2. Use `astria query`, `path` (add `--directed` for call direction), `explain`, or `affected` for cross-module questions
-3. Only use native file tools AFTER the graph identified the exact files
-
-## After editing code
-
-Run `astria update .` to keep the graph current.
+---
+name: astria
+description: Turn any directory into a queryable knowledge graph. Trigger: /astria
+---
+
+# astria skill (Kiro)
+
+When the user types `/astria`, run the astria knowledge graph pipeline.
+
+## Step 1 - Build or update the graph
+
+Run via terminal:
+```bash
+node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
+```
+
+- `missing` → run `astria run .`
+- `stale` → run `astria update .`
+- `fresh` → skip to Step 2
+
+## Step 2 - Read the report
+
+Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
+## Prefer MCP tools when connected
+
+If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
+
+
+## Enforcement Rules
+
+The `.kiro/steering/astria.md` steering file enforces graph usage. You MUST:
+
+1. Read `.astria/graph_report.md` before searching files
+2. Use `astria query`, `path` (add `--directed` for call direction), `explain`, or `affected` for cross-module questions
+3. Only use native file tools AFTER the graph identified the exact files
+
+## After editing code
+
+Run `astria update .` to keep the graph current.

package/skills/skill-opencode.md

@@ -1,44 +1,44 @@
----
-name: astria
-description: Turn any directory into a queryable knowledge graph. Trigger: /astria
----
-
-# astria skill (OpenCode)
-
-When the user types `/astria`, run the astria knowledge graph pipeline.
-
-## Step 1 - Build or update the graph
-
-Run via bash:
-```bash
-node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
-```
-
-- `missing` → run `astria run .`
-- `stale` → run `astria update .`
-- `fresh` → skip to Step 2
-
-## Step 2 - Read the report
-
-Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
-## Prefer MCP tools when connected
-
-If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
-
-
-## Enforcement Rules
-
-When `.astria/` exists, the astria plugin intercepts `view`, `grep`, `glob`, `ls`, and `bash` tools. You MUST use astria commands first:
-
-| Question | Command |
-|----------|---------|
-| Where is X? | `astria query "X"` |
-| How does X connect to Y? | `astria path "X" "Y"` |
-| What breaks if X changes? | `astria affected "X"` |
-| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
-| What does X do? | `astria explain "X"` |
-| Architecture overview? | Read `.astria/graph_report.md` |
-
-## After editing code
-
-Run `astria update .` to keep the graph current.
+---
+name: astria
+description: Turn any directory into a queryable knowledge graph. Trigger: /astria
+---
+
+# astria skill (OpenCode)
+
+When the user types `/astria`, run the astria knowledge graph pipeline.
+
+## Step 1 - Build or update the graph
+
+Run via bash:
+```bash
+node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
+```
+
+- `missing` → run `astria run .`
+- `stale` → run `astria update .`
+- `fresh` → skip to Step 2
+
+## Step 2 - Read the report
+
+Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
+## Prefer MCP tools when connected
+
+If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
+
+
+## Enforcement Rules
+
+When `.astria/` exists, the astria plugin intercepts `view`, `grep`, `glob`, `ls`, and `bash` tools. You MUST use astria commands first:
+
+| Question | Command |
+|----------|---------|
+| Where is X? | `astria query "X"` |
+| How does X connect to Y? | `astria path "X" "Y"` |
+| What breaks if X changes? | `astria affected "X"` |
+| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
+| What does X do? | `astria explain "X"` |
+| Architecture overview? | Read `.astria/graph_report.md` |
+
+## After editing code
+
+Run `astria update .` to keep the graph current.

package/skills/skill-trae.md

@@ -1,46 +1,46 @@
----
-name: astria
-description: Turn any directory into a queryable knowledge graph. Trigger: /astria
----
-
-# astria skill (Trae)
-
-When the user types `/astria`, run the astria knowledge graph pipeline.
-
-## Step 1 - Build or update the graph
-
-Run via terminal:
-```bash
-node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
-```
-
-- `missing` → run `astria run .`
-- `stale` → run `astria update .`
-- `fresh` → skip to Step 2
-
-## Step 2 - Read the report
-
-Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
-## Prefer MCP tools when connected
-
-If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
-
-
-## Enforcement Rules
-
-When `.astria/` exists, you MUST use astria commands BEFORE searching files:
-
-| Question | Command |
-|----------|---------|
-| Where is X? | `astria query "X"` |
-| How does X connect to Y? | `astria path "X" "Y"` |
-| What breaks if X changes? | `astria affected "X"` |
-| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
-| What does X do? | `astria explain "X"` |
-| Architecture overview? | Read `.astria/graph_report.md` |
-
-Only use native file tools AFTER the graph identified the exact files.
-
-## After editing code
-
-Run `astria update .` to keep the graph current.
+---
+name: astria
+description: Turn any directory into a queryable knowledge graph. Trigger: /astria
+---
+
+# astria skill (Trae)
+
+When the user types `/astria`, run the astria knowledge graph pipeline.
+
+## Step 1 - Build or update the graph
+
+Run via terminal:
+```bash
+node -e "const fs=require('fs');if(!fs.existsSync('.astria/graph.json')){console.log('missing')}else{const age=Math.round((Date.now()-fs.statSync('.astria/graph.json').mtimeMs)/60000);console.log(age>30?'stale':'fresh')}"
+```
+
+- `missing` → run `astria run .`
+- `stale` → run `astria update .`
+- `fresh` → skip to Step 2
+
+## Step 2 - Read the report
+
+Read `.astria/graph_report.md` and summarize: hub nodes, communities, surprising connections.
+## Prefer MCP tools when connected
+
+If the `astria` MCP server is connected (registered by `astria install` for claude, cursor, gemini, and zcode), prefer its native tools over shell commands: `repo_map` to orient, `query_graph` for a natural-language question, `explain`/`get_neighbors` for one symbol, `shortest_path` to trace a connection, `affected` before changing a shared symbol. Use the CLI commands below only when the server is not connected.
+
+
+## Enforcement Rules
+
+When `.astria/` exists, you MUST use astria commands BEFORE searching files:
+
+| Question | Command |
+|----------|---------|
+| Where is X? | `astria query "X"` |
+| How does X connect to Y? | `astria path "X" "Y"` |
+| What breaks if X changes? | `astria affected "X"` |
+| How does X reach Y (call direction)? | `astria path "X" "Y" --directed` |
+| What does X do? | `astria explain "X"` |
+| Architecture overview? | Read `.astria/graph_report.md` |
+
+Only use native file tools AFTER the graph identified the exact files.
+
+## After editing code
+
+Run `astria update .` to keep the graph current.