@miller-tech/uap 1.163.13 → 1.164.0

package/docs/guides/DROIDS_AND_SKILLS.md

@@ -1,6 +1,6 @@
 # Droids and Skills
 
-> Applies to UAP **v1.93.1**
+> Applies to UAP **v1.163**
 
 > **🏭 Where this fits:** PREP/ROUTING — this is the station where a
 > generalist agent takes the wrong approach on specialist work: reviewing

package/docs/guides/LOCAL_MODELS.md

@@ -1,6 +1,6 @@
 # Running UAP Against Local Models
 
-> UAP v1.93.1
+> UAP v1.163
 
 > **🏭 Where this fits:** BUILD — this is the station where a cheap local model
 > flails: plausible-but-wrong code, empty output, or a loop that never lands a

package/docs/guides/MCP_ROUTER.md

@@ -1,6 +1,6 @@
 # MCP Router
 
-> UAP v1.93.1
+> UAP v1.163
 
 > **🏭 Where this fits:** CROSS-CUTTING — keeping the context window lean at every station of the [delivery pipeline](./DELIVERY_PIPELINE.md). In a normal agentic workflow every tool call dumps its full result into context, so mostly-irrelevant output crowds out the details your agent actually needs. **What it delivers:** a proxy that compresses tool output before it reaches the model — up to 98% fewer tokens on large results — so the agent keeps room for real work and you spend less per session.
 

package/docs/guides/MEMORY.md

@@ -1,6 +1,6 @@
 # Memory System
 
-> UAP v1.93.1
+> UAP v1.163
 
 > **🏭 Where this fits:** INTAKE + FEEDBACK — the front and back doors of the [delivery pipeline](./DELIVERY_PIPELINE.md). At intake, a fresh agent forgets everything the last session learned and re-hallucinates scope; at feedback, nobody records the lesson, so the same mistake ships again next week. **What it delivers:** a persistent, searchable memory that hands each new agent the relevant past learnings on the way in, and captures durable lessons on the way out — so your team of agents gets smarter every run instead of starting from zero.
 
@@ -199,7 +199,7 @@ uap memory correct <search> [options]
 ```bash
 uap memory correct "uses pgvector" \
   --correction "uses Qdrant for semantic recall" \
-  --reason "migrated in v1.26"
+  --reason "superseded by newer approach"
 ```
 
 ### Maintain

package/docs/guides/MULTI_MODEL.md

@@ -1,6 +1,6 @@
 # Multi-Model Routing
 
-> Applies to UAP **v1.93.1**
+> Applies to UAP **v1.163**
 
 > **🏭 Where this fits:** PREP/ROUTING — this is the station where a normal
 > agentic workflow sends *everything* to one model: paying premium-model rates

package/docs/guides/POLICIES.md

@@ -1,6 +1,6 @@
 # Policies
 
-> Applies to UAP v1.93.1
+> Applies to UAP v1.163
 
 > **🏭 Where this fits:** CROSS-CUTTING — the executable rules bolted to every station of the [delivery pipeline](./DELIVERY_PIPELINE.md). In a normal agentic workflow the "rules" live in a prose prompt the model is free to ignore; that's how work escapes isolation, skips tests, or ships without review. **What it delivers:** each rule is a Python enforcer that actually inspects an operation and can *block* it before it runs — worktree isolation, test deltas, expert review, schema diffs, artifact hygiene, and more — so the guardrails hold instead of merely being suggested.
 

package/docs/guides/PROXY.md

@@ -51,16 +51,17 @@ the last session leaves — but it **never** kills a proxy that systemd manages
 that other sessions are still using.
 
 ```
-uap proxy [ensure | release | status | start | stop | restart | enable | disable]
+uap proxy [ensure | release | status | start | stop | restart | enable | disable | dashboard [on|off]]
 ```
 
 | Subcommand | What it does |
 |---|---|
-| `ensure` | Start the proxy if none is running, or **adopt** a running one; register this session as a client |
-| `release` | Deregister this session; stop the proxy **only if** we spawned it as a plain process and no other client remains |
-| `status` | Report whether it's running, how (systemd vs spawned), the port, and client count (`--json` for machine-readable) |
+| `ensure` | Start the proxy if none is running, or **adopt** a running one; register this session as a client; start-or-adopt the dashboard |
+| `release` | Deregister this session; stop the proxy (and dashboard) **only if** we spawned it as a plain process and no other client remains |
+| `status` | Report whether the proxy and dashboard are running, how (systemd vs spawned), ports, URL, and client count (`--json` for machine-readable) |
 | `start` / `stop` / `restart` | Explicit control (e.g. `restart` after changing `PROXY_*` in `.uap/proxy.env`) |
 | `enable` / `disable` | Toggle `.uap.json` `proxy.autostart` so hooks auto-start it (or not) |
+| `dashboard [on\|off]` | Show or toggle the ride-along dashboard (`.uap.json` `proxy.dashboard`) |
 
 | Flag | Purpose |
 |---|---|
@@ -68,8 +69,50 @@ uap proxy [ensure | release | status | start | stop | restart | enable | disable
 | `--client-pid <n>` | Long-lived agent pid for liveness (hooks pass `$PPID`) |
 | `--port <n>` | Proxy port (default `4000` / `$PROXY_PORT`) |
 | `--if-enabled` | No-op unless `.uap.json` `proxy.autostart` is `true` (hook-safe) |
+| `--no-dashboard` | Don't *start* the ride-along dashboard on this `ensure`/`start` (teardown still reaps a dashboard we own) |
 | `--quiet` / `--json` | Suppress output (hooks) / machine-readable status |
 
+### Ride-along dashboard
+
+**The operational dashboard starts with the proxy** — no second command. `ensure`
+and `start` also start-or-adopt `uap dashboard serve` (default
+<http://localhost:3847>), concurrently with the proxy so session start isn't
+serialized behind two health waits. Teardown follows ownership: a dashboard **we**
+spawned stops when the last client leaves (`release`) or immediately on an
+explicit `stop`; one you started yourself is adopted and **never** stopped. If a
+foreign process holds the port, UAP declines to spawn (fast, no stall) and tells
+you to run `uap dash serve` manually.
+
+| Setting | Where | Default |
+|---|---|---|
+| enabled | `.uap.json` `proxy.dashboard` (bool or `{enabled, port, host}`) / `UAP_PROXY_DASHBOARD=0\|1` | on |
+| port | `proxy.dashboard.port` / `UAP_DASH_PORT` | `3847` |
+| host | `proxy.dashboard.host` / `UAP_DASH_HOST` | `localhost` |
+| startup wait | `UAP_DASH_HEALTH_WAIT_MS` | `10000` |
+
+Env wins over config. These govern the **ride-along** only — `uap dash serve`
+still takes its own `--port`/`--host`.
+
+**Identification.** Liveness is `GET /health`, which returns
+`{"ok":true,"service":"uap-dashboard","port":…,"root":"<project dir>"}`. UAP adopts
+a dashboard only when that marker is present *and* `root` matches the current
+project — a dashboard is per-project (every panel reads its own working
+directory), even though the proxy client registry is per-user. A dashboard from an
+older UAP has no `/health`; it is recognised by its `UAP Dashboard` page title and
+adopted as-is, since it cannot state which project it serves.
+
+```bash
+uap proxy dashboard          # is it on, and where?
+uap proxy dashboard off      # opt this project out
+uap proxy start --no-dashboard
+```
+
+**Security note.** Because the dashboard is now up for the whole session rather
+than only while you run it by hand, the page carrying the policy-mutation token is
+served same-origin only, WebSocket upgrades require a same-origin `Origin`, and the
+token is not echoed into the ride-along log. Binding beyond loopback
+(`UAP_DASH_HOST`) is reported with a warning — set `UAP_DASHBOARD_TOKEN` if you do.
+
 **How it's managed.** When available, the proxy runs as the systemd unit
 `uap-anthropic-proxy.service`, reading its environment from
 `~/.config/uap/anthropic-proxy.env`; a plain spawned proxy seeds the same file

package/docs/guides/SANDBOX.md

@@ -65,7 +65,7 @@ it and calls `_strip_sandbox_unreachable_tools` (`anthropic_proxy.py:4393`), rem
 tool whose name starts with a prefix in `PROXY_SANDBOX_UNREACHABLE_PREFIXES` (default
 `mcp__claude-in-chrome__`). The model never sees the unreachable browser tools and falls
 back to `WebFetch` / local reads instead. (This is the "strip unreachable browser MCP tools
-from sandboxed sessions" change from v1.63.0.)
+from sandboxed sessions" behavior.)
 
 ## Escape hatches (operator-only)
 

package/docs/guides/WORKTREE_WORKFLOW.md

@@ -1,6 +1,6 @@
 # Worktree Workflow
 
-> Applies to UAP v1.93.1
+> Applies to UAP v1.163
 
 > **🏭 Where this fits:** ISOLATION — the station where a normal agentic workflow smears half-finished edits across `main`, clobbers files, and lets parallel agents collide into corrupt merge state. **What it delivers:** every agent works on its own branch in its own checkout, so the project root stays clean, each unit of work has a tidy PR boundary, and any number of agents can run at once without stepping on each other.
 

package/docs/reference/CLI.md

@@ -126,7 +126,7 @@ uap setup [options]
 | `--no-self-update` | Skip the automatic UAP CLI version check / self-update (also `UAP_NO_SELF_UPDATE=1`) |
 | `--systemd-services` | Scaffold user systemd services for llama.cpp + anthropic proxy |
 | `-d, --project-dir <path>` | Target project directory (defaults to cwd) |
-| `-i, --interactive` | Run the guided wizard (now the default; kept for back-compat) |
+| `-i, --interactive` | Run the guided wizard (the default) |
 
 The guided wizard (default) prompts for harnesses, memory tiers, coordination,
 patterns, policies, model provider/profile, hooks, and browser — with smart
@@ -534,7 +534,18 @@ Anthropic-compatible gateway in front of a local llama.cpp/Qwen). Hooks
 session leaves — but it never kills a proxy that systemd manages or that other
 sessions still use.
 
-`uap proxy [ensure | release | status | start | stop | restart | enable | disable]`
+`uap proxy [ensure | release | status | start | stop | restart | enable | disable | dashboard [on|off]]`
+
+The **operational dashboard rides along**: `ensure`/`start` also start-or-adopt
+`uap dashboard serve` (default <http://localhost:3847>), so monitoring never needs
+a second command. `release` stops it when the last client leaves; `stop` shuts it
+down immediately (ownership only — it does not consult the client count). A
+dashboard you started yourself, or one serving a different project, is never
+touched. Opt out per project with `uap proxy dashboard off`, per `ensure`/`start`
+with `--no-dashboard`, or globally with `UAP_PROXY_DASHBOARD=0`.
+Ports/hosts: `UAP_DASH_PORT` (3847), `UAP_DASH_HOST` (localhost),
+`UAP_DASH_HEALTH_WAIT_MS` (10000) — these govern the ride-along, not
+`uap dash serve`.
 
 | Flag | Purpose |
 |------|---------|
@@ -542,11 +553,13 @@ sessions still use.
 | `--client-pid <n>` | Long-lived agent pid for liveness (hooks pass `$PPID`) |
 | `--port <n>` | Proxy port (default 4000 / `$PROXY_PORT`) |
 | `--if-enabled` | No-op unless `.uap.json` `proxy.autostart` is true (hook-safe) |
+| `--no-dashboard` | Don't start the ride-along dashboard on this `ensure`/`start` |
 | `--quiet` / `--json` | Suppress output (hooks) / machine-readable status |
 
 ```bash
-uap proxy status --json
+uap proxy status --json    # includes a `dashboard` block (port, url, healthy, owner)
 uap proxy restart          # e.g. after changing PROXY_* in .uap/proxy.env
+uap proxy dashboard off    # opt out of the ride-along dashboard
 ```
 
 The proxy binds `127.0.0.1` by default; see the proxy `PROXY_*` settings in the

package/docs/reference/CONFIGURATION.md

@@ -1,6 +1,6 @@
 # Configuration Reference
 
-> Universal Agent Protocol (UAP) v1.93.1
+> Universal Agent Protocol (UAP) v1.163
 
 > **🏭 Where this fits:** Cross-cutting — the settings that tune every station.
 > **What it delivers:** the dials that decide how your [delivery pipeline](../guides/DELIVERY_PIPELINE.md)

package/docs/reference/FEATURES.md

@@ -1,6 +1,6 @@
 # Feature Catalog
 
-> Universal Agent Protocol (UAP) v1.93.1
+> Universal Agent Protocol (UAP) v1.163
 
 > **🏭 Where this fits:** Every station on the line — this catalog is the parts list for the whole factory. **What it delivers:** each feature below is mapped to the stage of your [delivery pipeline](../guides/DELIVERY_PIPELINE.md) it protects, so you can see exactly where your agent's work stops being plausible-looking and starts being genuinely shippable.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@miller-tech/uap",
-  "version": "1.163.13",
+  "version": "1.164.0",
   "description": "Autonomous AI agent memory system with CLAUDE.md protocol enforcement",
   "type": "module",
   "main": "dist/index.js",

package/templates/hooks/session-end.sh

@@ -58,4 +58,13 @@ if [ -d "$BACKUP_DIR" ]; then
   find "$BACKUP_DIR" -maxdepth 1 -mindepth 1 -type d -mtime +7 -exec rm -rf {} \; 2>/dev/null || true
 fi
 
+# Session-scoped proxy + dashboard release: stop them ONLY if THIS session
+# started them and no other client remains (adopted/externally-managed services
+# are never killed; other active clients keep them alive). Fail-open; never
+# blocks exit.
+if command -v uap >/dev/null 2>&1; then
+  _uap_pc="${CLAUDE_SESSION_ID:-${FACTORY_SESSION_ID:-${CURSOR_SESSION_ID:-${UAP_SESSION_ID:-ppid-$PPID}}}}"
+  ( cd "$PROJECT_DIR" 2>/dev/null && timeout 20 uap proxy release --if-enabled --quiet --client "$_uap_pc" --client-pid "$PPID" ) >/dev/null 2>&1 || true
+fi
+
 exit 0

package/templates/hooks/session-start.sh

@@ -250,6 +250,16 @@ if command -v uap >/dev/null 2>&1; then
   fi
 fi
 
+# Session-scoped proxy autostart (opt-in via .uap.json proxy.autostart). Starts
+# the proxy if none is running, or ADOPTS an already-running one, and brings the
+# operational dashboard up with it (http://localhost:3847 — opt out with
+# `uap proxy dashboard off`). Reference-counted: released only when the last
+# client leaves. Fail-open, never blocks.
+if command -v uap >/dev/null 2>&1; then
+  _uap_pc="${CLAUDE_SESSION_ID:-${FACTORY_SESSION_ID:-${CURSOR_SESSION_ID:-${UAP_SESSION_ID:-ppid-$PPID}}}}"
+  ( cd "$PROJECT_DIR" 2>/dev/null && timeout 30 uap proxy ensure --if-enabled --quiet --client "$_uap_pc" --client-pid "$PPID" ) >/dev/null 2>&1 || true
+fi
+
 if [ -n "$output" ]; then
   echo "$output"
 fi