xtrm-tools 0.5.45 → 0.5.47

package/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "xtrm-tools",
-  "version": "0.5.45",
+  "version": "0.5.47",
   "description": "xtrm-tools: dual-runtime workflow enforcement (Claude Code + Pi) — hooks, extensions, skills, and MCP servers",
   "author": {
     "name": "jaggers"

package/CHANGELOG.md

@@ -7,6 +7,20 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ---
 
+## [0.5.45] - 2026-03-25
+
+### Changed
+- **`xt memory update`**: Replaced raw specialist stream with ora spinner + final summary output. Shows animated spinner while specialist runs; on finish prints `✓ .xtrm/memory.md written.` (or `✗`) followed by the last 10 meaningful lines dimmed.
+
+---
+
+## [0.5.44] - 2026-03-25
+
+### Added
+- **`xt help`**: `xtrm memory update` entry added to PRIMARY COMMANDS section.
+
+---
+
 ## [0.5.43] - 2026-03-25
 
 ### Fixed

package/README.md

@@ -20,7 +20,7 @@ The `planning` skill generates a structured issue board from any spec or idea
 
 ### Statusline
 
-A live statusline renders in every Claude Code session: active claim, open issue count, model, context window health (color-coded truecolor gradient), and token usage — all in a single line below the prompt. No configuration required after `xtrm install`.
+A live statusline renders in every Claude Code session: active claim, open issue count, model, context window health (color-coded truecolor gradient), and token usage — all in a single line below the prompt. No configuration required after `xtrm init`.
 
 ### Specialists *(upcoming)*
 
@@ -57,8 +57,8 @@ Native integration with the [specialists](https://github.com/Jaggerxtrm/speciali
 # Install globally (one-time)
 npm install -g github:Jaggerxtrm/xtrm-tools@latest
 
-# Install the plugin
-xtrm install
+# Set up xtrm in your project
+xtrm init
 
 # Verify
 claude plugin list
@@ -67,9 +67,26 @@ claude plugin list
 
 **One-line run:**
 ```bash
-npx -y github:Jaggerxtrm/xtrm-tools install
+npx -y github:Jaggerxtrm/xtrm-tools init
 ```
 
+**Typical workflow after install:**
+```bash
+# Start a sandboxed session in a worktree
+xt claude my-feature
+
+# Publish that worktree: rebase, push, open PR, optional cleanup
+xt end
+
+# Refresh project memory from bd memories + current repo state
+xt memory update
+
+# If multiple xt/* PRs are open, drain the merge queue oldest-first
+xt merge
+```
+
+`xt end` handles one worktree session at a time. `xt merge` is the follow-up queue operator: it inspects open `xt/*` PRs, processes them FIFO, waits for green CI on the oldest PR, merges it with `--rebase`, then rebases the remaining queued xt branches and repeats. `xt memory update` shells out to the `memory-processor` specialist, which condenses bd memories and current project state into `.xtrm/memory.md`; use `--dry-run` to inspect without writing.
+
 ---
 
 ## What's Included
@@ -176,6 +193,8 @@ xtrm <command> [options]
 | `pi [name]` | Launch Pi in a sandboxed `xt/<name>` worktree |
 | `attach [slug]` | Re-attach to an existing worktree and resume the Claude or Pi session |
 | `end` | Close worktree session: rebase, push, PR, cleanup |
+| `memory update` | Run `memory-processor` to synthesize bd memories + repo state into `.xtrm/memory.md` |
+| `merge` | Drain queued `xt/*` PRs via `xt-merge`: FIFO CI gate → rebase merge → rebase cascade |
 | `worktree list` | List active `xt/*` worktrees with runtime, last activity, and resume hint |
 | `worktree clean` | Remove merged worktrees |
 | `docs` | Documentation inspection and drift-check suite (`xtrm docs --help`) |
@@ -200,7 +219,7 @@ See [docs/cli-architecture.md](docs/cli-architecture.md) for internals.
 | `github-grep` | Code search |
 | `deepwiki` | Repository documentation |
 
-Official Claude plugins installed by `xtrm install`: `serena`, `context7`, `github`, `ralph-loop`.
+Official Claude plugins installed by `xtrm init`: `serena`, `context7`, `github`, `ralph-loop`.
 
 See [docs/mcp-servers.md](docs/mcp-servers.md) for configuration details.