v0.2.0: probe install + refresh-before-search

.gitignore

@@ -4,3 +4,4 @@ __pycache__/
 *.egg-info/
 dist/
 build/
+.worktrees/

CHANGELOG.md

@@ -0,0 +1,17 @@
+# Changelog
+
+All notable changes to probe are documented here.
+
+## 0.2.0 — 2026-04-17
+
+### Added
+- `probe install` / `probe uninstall` for one-command Claude Code integration via `claude mcp add --scope user`.
+- Refresh-before-search: the index incrementally updates when files change, driven by a two-phase (stat → hash) algorithm with a TTL debounce (`PROBE_REFRESH_TTL`, default 5s).
+- `refreshed` field on every `probe_search` MCP response with counts of added/changed/removed files.
+
+### Changed
+- MCP `probe_search` no longer needs a separate `probe_index` call — auto-index is subsumed into the unified refresh path.
+- `files` table schema gains `mtime_ns` and `size` columns (migration is automatic and backwards-compatible).
+
+### Removed
+- The "File system watcher" roadmap item (refresh-before-search covers the same UX need without a daemon).

README.md

@@ -21,25 +21,31 @@ probe makes this instant. It indexes everything -- markdown docs, code, PDFs, pl
 # 2. Install
 pip install probe-search
 
-# 3. Set your API key
-export ZEROENTROPY_API_KEY="ze_xxx"
+# 3. Register probe with Claude Code (one-time, machine-wide)
+probe install
 
-# 4. Index and search
-probe index .
-probe search "how does authentication work"
+# Now open any project in Claude Code and ask a question —
+# probe will auto-index on first search and refresh on subsequent ones.
 ```
 
-Or run without installing:
+For CLI-only use:
 
 ```bash
-uvx probe-search search "how does authentication work"
+export ZEROENTROPY_API_KEY="ze_xxx"
+probe index .
+probe search "how does authentication work"
 ```
 
+The index auto-refreshes before each search; set `PROBE_REFRESH_TTL=0` to
+force refresh every time, or `-1` to disable refresh.
+
 ---
 
 ## MCP Server Setup (Claude Code, Cursor)
 
-Add a `.mcp.json` file to your project root:
+**Claude Code users**: `probe install` (see Quick Start) does this automatically.
+
+**For Cursor or advanced use**: add a `.mcp.json` file to your project root:
 
 ```json
 {
@@ -120,6 +126,7 @@ One query returns the design spec, the implementation code, and the architectura
 |---------|-------------|
 | `probe index [paths...]` | Index project files for semantic search |
 | `probe index --full` | Force full re-index (ignore file hashes) |
+| `probe install` | Register probe as a user-scope MCP server in Claude Code |
 | `probe search "query"` | Search project knowledge with natural language |
 | `probe search --top-k N` | Limit number of results (default: 10) |
 | `probe search --type code` | Filter by file type (markdown, code, pdf, text) |
@@ -130,6 +137,7 @@ One query returns the design spec, the implementation code, and the architectura
 | `probe config` | Show current provider configuration |
 | `probe init` | Auto-detect provider and save config |
 | `probe mcp` | Start MCP server (stdio transport) |
+| `probe uninstall [--purge]` | Unregister probe; `--purge` also deletes `.probe/` in cwd |
 
 ---
 
@@ -181,7 +189,7 @@ Documents are chunked and stored locally in `.probe/` (SQLite + numpy). Only chu
 
 ## What's NOT in v1
 
-- File system watcher for auto-reindexing on changes
+- Real-time filesystem watcher (refresh-before-search handles typical edit volumes fine)
 - Web sources (Notion, Confluence, Google Docs)
 - Git-aware context (commit history, blame)
 - Image/diagram understanding within PDFs