Enforce chat-only-on-explicit-request policy across all tool surfaces

- chat.py: remove promotional capabilities list, replace soft Note with
  hard "disabled by default" rule
- codealive_mcp_server.py: rewrite instructions — separate default workflow
  from user-invoked chat, remove "after search" justification
- codebase_consultant: add explicit invocation policy matching chat
- Bump version to 2.0.1

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: rodion-m

src/codealive_mcp_server.py

@@ -52,27 +52,32 @@
     - Answer questions about code implementation details
     - Integrate with local git repositories for seamless code exploration
 
-    When working with a codebase, follow this workflow:
+    Default workflow (used for ALL tasks unless the user explicitly requests `chat`):
     1. First use `get_data_sources` to identify available repositories and workspaces
     2. Use `semantic_search` for natural-language retrieval by meaning
     3. Use `grep_search` for literal string or regex matching when the pattern matters
-    3. To get full content:
+    4. To get full content:
        - For repos in your working directory: use `Read()` on the local files
        - For external repos: use `fetch_artifacts` with identifiers from search results
-    4. Use `get_artifact_relationships` or `fetch_artifacts` to drill into the most relevant hits
-    5. If your environment supports subagents and you need the highest reliability or depth,
+    5. Use `get_artifact_relationships` or `fetch_artifacts` to drill into the most relevant hits
+    6. If your environment supports subagents and you need the highest reliability or depth,
        prefer an agentic workflow where a subagent combines `semantic_search`, `grep_search`,
        artifact fetches, relationship inspection, and local file reads
-    6. Use `chat` only when you specifically need a synthesized answer after search;
-       it is usually not the default choice and can take up to 30 seconds
+
+    User-invoked tool — `chat`:
+    - `chat` is disabled by default. Call it ONLY when the user has explicitly
+      named the tool (e.g. "use chat", "ask CodeAlive", "use codebase_consultant").
+    - For every other case — lookups, architecture understanding, debugging,
+      summaries — use semantic_search, grep_search, fetch_artifacts, and
+      get_artifact_relationships. Do not treat "after search" as a justification
+      for calling chat.
 
     For effective code exploration:
     - Start with broad natural-language queries in `semantic_search` to understand the overall structure
     - Use `grep_search(regex=false)` for exact strings and `grep_search(regex=true)` for regex patterns
     - Use specific function/class names or file path scopes when looking for particular implementations
     - Treat `semantic_search` and `grep_search` as the default discovery tools
     - Prefer `semantic_search` over the deprecated `codebase_search` legacy alias
-    - Reserve `chat` for synthesis after search, not for first-pass evidence gathering
     - Remember that context from previous messages is maintained in the same conversation
 
     Flexible data source usage:

src/tools/chat.py

@@ -43,12 +43,10 @@ async def chat(
     UNLESS the user has explicitly provided specific data source names OR you are continuing an
     existing conversation with a `conversation_id`.
 
-    This tool understands the indexed codebase and can help with:
-    - Architecture and design decisions
-    - Implementation strategies
-    - Code explanations and walkthroughs
-    - Best practices and optimization advice
-    - Debugging and problem-solving
+    When invoked by the user, this tool can produce synthesized answers about
+    architecture, design decisions, code walkthroughs, debugging, etc.
+    These topics do NOT by themselves justify calling the tool — only an
+    explicit user request does.
 
     Args:
         question: What you want to know about the codebase
@@ -84,8 +82,8 @@ async def chat(
            )
 
     Note:
-        - `chat` is usually not needed for simple lookups or evidence gathering
-        - Prefer `semantic_search` and `grep_search` as the default tools
+        - `chat` is disabled by default; see top of docstring for the single
+          condition that permits calling it.
         - Either conversation_id OR data_sources is typically provided
         - When creating a new conversation, data_sources is optional if your API key has exactly one assigned data source
         - When continuing a conversation, conversation_id is required to maintain context
@@ -110,8 +108,12 @@ async def codebase_consultant(
     """Deprecated alias for `chat`.
 
     Keep this for backward compatibility with older prompts and MCP clients.
-    New integrations should prefer `chat`, while default discovery should still
-    start with `semantic_search` and `grep_search`.
+    New integrations should prefer the canonical `chat` tool name.
+
+    **Same invocation policy as `chat`**: do NOT call unless the user explicitly
+    named the tool (e.g. "use codebase_consultant", "ask CodeAlive", "use chat").
+    For all other tasks use `semantic_search`, `grep_search`, `fetch_artifacts`,
+    and `get_artifact_relationships`.
     """
     return await _chat_impl(
         ctx,