context-mcp-server 1.0.2 → 1.0.4

package/src/templates/CLAUDE.md

@@ -9,7 +9,7 @@ description: >
 # Context-MCP — Claude Usage Guide
 
 Persistent memory + codebase knowledge graph for Claude.
-Every conversation starts with `context.resume`. Every structural question uses `codegraph_query`. Files only read for bugs/logic.
+Every conversation starts with `context.resume`. Every codebase question uses `codegraph_query`. Files only read for bugs/logic.
 
 ---
 
@@ -45,42 +45,31 @@ Always pass `project`. Auto-compact fires at >50 entries — oldest summarized a
 
 ## 3. CodeGraph Pipeline
 
-### Step 1 — Build (once per project, free)
+### Step 1 — Build (once per project, fast, local)
 ```
 codegraph_build(path)
-  → AST: code files → functions, classes, imports, edges
-  → Config: .yaml .toml .sql → schema nodes
-  → Docs: .md .txt .pdf → pending (no content yet)
 ```
-Saves graph to `~/.context-mcp/graphs.json`. Visible on `context.resume`.
+- Parses codebase into AST graph using tree-sitter (regex fallback for unsupported languages)
+- Extracts functions, classes, imports, call edges for all code files
+- Build files (package.json, pyproject.toml, go.mod, Dockerfile, etc.) get a single metadata node
+- Saves graph to `~/.context-mcp/graphs.json`. Visible on `context.resume`.
 
-### Step 2 — Enrich (one-time cost per file)
+### Step 2 — Query (free, instant, forever)
 ```
-codegraph_extract(path)
-  → returns changed code files (with existing node list) + doc files (raw text)
-
-For each code file: write description for each node listed in existing_nodes
-For each doc file:  extract concept nodes + relationships
-
-codegraph_add_nodes(path, nodes)
-  → stores descriptions in semantic cache (never overwritten by rebuild)
-```
-
-### Step 3 — Query (free, instant forever)
-```
-codegraph_query(path, question)   → NODE/EDGE subgraph, token_budget param (default 2000)
-codegraph_explain(path, node)     → one node: description + depends_on + used_by
-codegraph_path(path, from, to)    → shortest path between two concepts
-codegraph_nodes(path, type)       → list all nodes of a type
-codegraph_report(path)            → god nodes, clusters, surprises
+codegraph_query(path, question)     → fetch any details about the codebase
+codegraph_explain(path, node)       → one node: type, file, depends_on, used_by
+codegraph_path(path, from, to)      → shortest path between two concepts
+codegraph_nodes(path, type)         → list all nodes of a type
+codegraph_report(path)              → god nodes, clusters, surprises
 ```
 
 ---
 
 ## 4. Graph vs File
 
-**Graph** — structural questions: dependencies, callers, imports, paths between concepts.
-**File** — bugs, logic inside a function, tracing unexpected behavior.
+**Graph** — use for any question about what exists in the codebase: finding functions, classes, files, understanding what a module contains, dependencies, callers, imports, paths between concepts.
+
+**File** — use for bugs, logic inside a specific function, tracing unexpected behavior.
 
 ---
 
@@ -91,4 +80,3 @@ codegraph_report(path)            → god nodes, clusters, surprises
 3. **`search` before asking** — if user references past work, find it first
 4. **`codegraph_query` before reading files** — graph is faster and cheaper
 5. **Read files for bugs/logic** — graph is structure only, not behavior
-6. **Enrich once** — run extract → add_nodes once per project; descriptions persist forever

package/src/templates/GEMINI.md

@@ -1,7 +1,7 @@
 # Context-MCP — Gemini CLI Usage Guide
 
 Persistent memory + codebase knowledge graph.
-Every conversation starts with `context.resume`. Every structural question uses `codegraph_query`. Files only read for bugs/logic.
+Every conversation starts with `context.resume`. Every codebase question uses `codegraph_query`. Files only read for bugs/logic.
 
 ---
 
@@ -37,21 +37,15 @@ Always pass `project`. Auto-compact fires at >50 entries.
 
 ## 3. CodeGraph Pipeline
 
-### Step 1 — Build (once, free)
+### Step 1 — Build (once, fast, local)
 ```
 codegraph_build(path)  →  AST graph: functions, classes, imports, edges
 ```
 
-### Step 2 — Enrich (one-time per file)
+### Step 2 — Query (free, instant)
 ```
-codegraph_extract(path)           →  file content + node list
-codegraph_add_nodes(path, nodes)  →  semantic descriptions (permanent cache)
-```
-
-### Step 3 — Query (free, instant)
-```
-codegraph_query(path, question)   →  NODE/EDGE subgraph (token_budget default 2000)
-codegraph_explain(path, node)     →  single node + neighbors
+codegraph_query(path, question)   →  fetch any details about the codebase
+codegraph_explain(path, node)     →  single node: type, file, connections
 codegraph_path(path, from, to)    →  shortest path
 codegraph_nodes(path, type)       →  list nodes by type
 codegraph_report(path)            →  full graph analysis
@@ -61,8 +55,8 @@ codegraph_report(path)            →  full graph analysis
 
 ## 4. Graph vs File
 
-**Graph** — structural questions: dependencies, callers, imports.
-**File** — bugs, logic, tracing behavior.
+**Graph** — use for any question about what exists: finding functions, classes, files, dependencies, callers, imports, paths between concepts.
+**File** — bugs, logic, tracing unexpected behavior.
 
 ---
 
@@ -73,4 +67,3 @@ codegraph_report(path)            →  full graph analysis
 3. **`search` before asking** — if user references past work, find it first
 4. **`codegraph_query` before reading files** — graph is faster and cheaper
 5. **Read files for bugs/logic** — graph is structure only, not behavior
-6. **Enrich once** — descriptions persist forever