@manny-est/node-red-flowpilot 0.5.1 → 0.6.0-beta.1

package/CHANGELOG.md

@@ -2,6 +2,64 @@
 
 All notable changes to FlowPilot are documented here.
 
+## [0.6.0-beta.1] - 2026-08-31
+
+Prerelease — published to the `beta` npm tag only. `latest` stays on `0.5.2`.
+
+### Added
+- **Agentic WRITE-tool loop for Modify** (behind the `enableAgentWrite`
+  setting, default off): step-by-step tool calls against the live flow
+  (`apply_step`, `remove_step`, `rename_node`, `group_nodes`), each
+  write-gated by a per-step consent prompt (Proceed / Skip this step) before
+  it touches the canvas. Multi-item requests are tracked and executed as
+  separate, individually-verified steps rather than one all-or-nothing
+  envelope; an injected mid-run failure (e.g. a referenced node no longer
+  exists) fails only that item, with the rest completing normally.
+- **`ask_user` clarifying-question tool**: an agent-strategy turn can pause
+  mid-run to ask a single focused question (with optional quick-reply
+  buttons) instead of guessing, and resumes exactly where it left off once
+  answered.
+- **Contract-exclusivity enforcement**: a server-side safety net
+  (`enforceAgentContract`) strips any classic-style mutation fields
+  (`changes`/`newNodes`/`newWires`/`removeNodes`/`newGroups`) that
+  accidentally appear on an agent-strategy turn with no tool calls, before
+  they ever reach the client — the two mutation paths (classic envelope vs.
+  agentic WRITE tools) stay mutually exclusive per turn.
+- **Run identity and honest interruption**: agent-strategy runs now carry a
+  stable run/operation id so a duplicate tool-call delivery (a retry, or the
+  model repeating itself) is applied at most once; a run that's stopped or
+  hits its step/token ceiling reports an honest "interrupted after N steps"
+  instead of silently truncating.
+
+### Security
+- **API keys are now write-only over HTTP** (previously exposed on
+  `GET`/`POST /flowpilot/settings` — a fork-hygiene regression from the
+  0.5.1 stable line, now ported forward and closed for good). Every
+  provider's `apiKey` is masked to a sentinel or `""` in both responses;
+  the real key never leaves the server. `settings.json` and per-conversation
+  transcripts are now created with `0600` permissions.
+- **Provider-confirmation gate (SSRF mitigation)**: no chat, generate,
+  modify, document, build, agent-step, or model-list request reaches a
+  configured provider's Base URL until that exact URL has passed a real
+  FlowPilot connection check (Pre-flight check / Test Provider). The check
+  itself is blind on failure — a non-provider target's response is never
+  reflected back to the client — so pointing a provider at an unintended
+  internal address yields nothing readable. Confirmation is tied to the
+  exact URL and clears automatically if the Base URL or API key changes.
+  See `dev-docs/decisions/ADR-007-Provider-Confirmation-Gate.md` for the
+  full design.
+- **Audit-trail completeness**: an agent-strategy request whose very first
+  model turn is a tool call (rather than a later continuation) is now
+  always recorded to the audit log — previously this specific case left no
+  trace at all.
+
+### Internal
+- Phase 10: strategy propagation, contract-exclusivity enforcement, agent
+  turn output caps, run identity/idempotency, and verification consolidation
+  across the classic and agentic Modify paths. Full history in
+  `dev-docs/current/Phase10-Build-Progress.md` and
+  `dev-docs/current/Phase10-Gate-Closeout-Final.md`.
+
 ## [0.5.1] - 2026-07-24
 
 ### Added

package/README.md

@@ -6,9 +6,10 @@ FlowPilot: AI assistance for Node-RED, designed for builders who want help
 without giving up control.
 
 FlowPilot is an AI-powered development assistant that lives in the Node-RED
-editor sidebar. It talks to any OpenAI-compatible API (OpenAI, LocalAI,
-Ollama, etc.) and helps you generate, modify, document, and discuss your
-flows — without ever acting behind your back.
+editor sidebar. It talks natively to Anthropic (Claude models) or to any
+OpenAI-compatible API (OpenAI, LocalAI, Ollama, etc.) and helps you
+generate, modify, document, and discuss your flows — without ever acting
+behind your back.
 
 ![FlowPilot sidebar](https://github.com/manny-est/flowpilot/releases/download/v0.2.1/sidebar-chat-overview.png)
 
@@ -29,8 +30,8 @@ installation, the sidebar UI, and a chapter on every feature.
 - **Undo first** — every change goes through Node-RED's native undo
   (Ctrl+Z), including multi-part changes (insertions + rewires + new nodes)
   as a single step.
-- **Open architecture** — provider-agnostic, OpenAI-compatible REST. No
-  lock-in to one AI platform.
+- **Open architecture** — provider-agnostic: native Anthropic support, or
+  any OpenAI-compatible REST endpoint. No lock-in to one AI platform.
 - **Simple and lightweight** — favors simple, maintainable solutions over
   speculative complexity.
 
@@ -129,11 +130,24 @@ under `<node-red-userDir>/flowpilot/`:
 - `chats/` — lightweight per-session chat logs
 - `backups/` — pre-change backups
 
-## Provider setup (example: LocalAI)
+## Provider setup
 
 Open the FlowPilot sidebar, click the settings (gear) icon, and add a
-provider:
+provider.
 
+**Anthropic (Claude models):**
+
+- Provider type: `Anthropic`
+- Provider name: `Claude` (or any label)
+- Base URL: leave blank (uses `api.anthropic.com`)
+- API key: your Anthropic API key. If it returns an
+  `anthropic-workspace-id is required` error, create a new key in the
+  Anthropic Console scoped to a single workspace.
+- Model: e.g. `claude-opus-5`, or click **Refresh models**
+
+**Example: LocalAI (OpenAI-compatible):**
+
+- Provider type: `OpenAI-compatible`
 - Provider name: `LocalAI` (or any label)
 - Base URL: `http://localhost:8080`
 - API key: blank unless your instance requires one

package/USER-GUIDE.md

@@ -103,29 +103,42 @@ saved transcript permanently.
 
 ## Set a Provider
 
-FlowPilot talks to any **OpenAI-compatible** API — OpenAI itself, LocalAI,
-Ollama (with its OpenAI-compatible endpoint), LM Studio, etc.
+FlowPilot talks to **Anthropic** (Claude models) natively, or to any
+**OpenAI-compatible** API — OpenAI itself, LocalAI, Ollama (with its
+OpenAI-compatible endpoint), LM Studio, etc.
 
 1. Open **Settings** (gear icon).
 2. Under **Providers**, click **+ Add** if you need a new provider slot
    (one is created for you by default).
-3. Fill in:
-   - **Provider Name** — any label, e.g. `LocalAI` or `OpenAI`.
-   - **Base URL** — e.g. `http://localhost:8080` or `https://api.openai.com`.
-     If Node-RED is running in Docker, `localhost` refers to the *Node-RED
-     container*, not the Docker host — use the provider's container name, a
-     Docker network alias, or a host IP (e.g. `http://172.17.0.1:8080`)
-     instead.
-   - **API Key** — leave blank unless your provider requires one.
-   - **Model** — type a model name, or click **Refresh models** to fetch the
-     provider's available models (via `GET /v1/models`) and pick from the
-     list.
+3. Pick a **Provider Type**: `OpenAI-compatible` (the default) or
+   `Anthropic`.
+4. Fill in:
+   - **Provider Name** — any label, e.g. `LocalAI`, `OpenAI`, or `Claude`.
+   - **Base URL** —
+     - OpenAI-compatible: e.g. `http://localhost:8080` or
+       `https://api.openai.com`. If Node-RED is running in Docker,
+       `localhost` refers to the *Node-RED container*, not the Docker
+       host — use the provider's container name, a Docker network alias,
+       or a host IP (e.g. `http://172.17.0.1:8080`) instead.
+     - Anthropic: leave blank to use `api.anthropic.com` — only set this
+       if you're routing through a proxy or gateway.
+   - **API Key** — for OpenAI-compatible, leave blank unless your provider
+     requires one. For Anthropic, this is required. If your key returns
+     an `anthropic-workspace-id is required` error, it wasn't scoped to a
+     single workspace when created — go to the Anthropic Console, create a
+     new key, and choose a specific workspace at creation time.
+   - **Model** — type a model name, or click **Refresh models** to fetch
+     the provider's available models and pick from the list.
    - **Temperature** — a starting value of `0.2` works well for most uses.
 
 ![Provider settings: Base URL, API key, model, and temperature fields](https://github.com/manny-est/flowpilot/releases/download/v0.2.1/sidebar-settings-providers.png)
 
 4. Click **Pre-flight check**. This saves your settings and sends a small
-   test request. A reply in the chat panel means you're connected.
+   test request. A reply in the chat panel means you're connected. **This
+   step is required** — a provider must pass Pre-flight check before Chat,
+   Generate, Modify, Document, or Build will send it anything; a provider
+   you just added or just changed the Base URL/API key on starts
+   unconfirmed until this check passes again.
 
 ![A successful Pre-flight check reply in Chat](https://github.com/manny-est/flowpilot/releases/download/v0.2.1/preflight-success.png)
 

package/flowpilot-core.css

@@ -15,6 +15,7 @@
 
 .fp-header-row {
     display: flex;
+    flex-wrap: wrap;
     align-items: center;
     gap: 12px;
 }
@@ -49,8 +50,12 @@
 .fp-view-buttons {
     display: flex;
     gap: 6px;
-    position: relative;
-    top: 2px;
+    /* Own row below the logo/title/subtitle at default (narrow) sidebar
+       width — sharing the row with .fp-heading left too little space for
+       "AI flow assistant" to fit on one line, so it wrapped across three. */
+    flex: 1 0 100%;
+    justify-content: flex-end;
+    margin-top: 8px;
 }
 
 .fp-panel {
@@ -244,7 +249,7 @@
     padding: 4px 0;
 }
 
-.fp-typing span:not(.fp-typing-label) {
+.fp-typing span:not(.fp-typing-label):not(.fp-typing-elapsed) {
     width: 7px;
     height: 7px;
     border-radius: 50%;
@@ -253,8 +258,15 @@
     animation: fp-bounce 1.2s infinite ease-in-out;
 }
 
-.fp-typing span:nth-child(2) { animation-delay: 0.18s; }
-.fp-typing span:nth-child(3) { animation-delay: 0.36s; }
+/* CLAUDE-033: matches the base rule's :not() clauses too, so specificity
+   (0,4,1) beats the base rule's (0,3,1) — otherwise the base rule's
+   "animation" SHORTHAND implicitly resets animation-delay back to 0s on
+   every dot regardless of source order, since shorthand properties reset
+   unspecified sub-values, and CSS resolves same-property conflicts by
+   specificity first. Without this, all three dots silently bounce in
+   perfect sync instead of staggered. */
+.fp-typing span:not(.fp-typing-label):not(.fp-typing-elapsed):nth-child(2) { animation-delay: 0.18s; }
+.fp-typing span:not(.fp-typing-label):not(.fp-typing-elapsed):nth-child(3) { animation-delay: 0.36s; }
 
 .fp-typing-label {
     font-size: 11px;
@@ -263,6 +275,12 @@
     align-self: center;
 }
 
+.fp-typing-elapsed {
+    font-size: 11px;
+    color: var(--red-ui-secondary-text-color, #999);
+    align-self: center;
+}
+
 .fp-agent-stop {
     margin-left: 8px;
     align-self: center;
@@ -366,6 +384,85 @@
     font-size: 16px;
 }
 
+/* Secondary/escape-hatch variant: transparent background, muted text.
+   Same hover glow as fp-chip-card so it reads as part of the same family
+   without competing visually with the primary action above it. */
+.fp-chip-card.fp-chip-card-alt {
+    background: transparent;
+    border-color: #2a333c;
+    color: #8a96a3;
+}
+.fp-chip-card.fp-chip-card-alt .fp-chip-title {
+    color: #8a96a3;
+    font-weight: 500;
+}
+.fp-chip-card.fp-chip-card-alt:hover {
+    /* CLAUDE-031: this variant's background stays transparent at rest by
+       design, so on Node-RED's LIGHT theme it inherits a white/near-white
+       surrounding background — the near-white hover text below then had
+       nothing dark to sit on and became unreadable. A solid dark fill on
+       hover (paired with the same green glow as the primary chip) fixes
+       contrast regardless of the surrounding theme, without touching the
+       transparent, theme-following look at rest. */
+    background: #141a21;
+    border-color: #46d39a;
+    box-shadow: inset 0 1px 0 rgba(255, 255, 255, 0.05), 0 0 0 1px #46d39a, 0 0 14px -2px #46d39a;
+    color: #dfe6ee;
+}
+.fp-chip-card.fp-chip-card-alt:hover .fp-chip-title {
+    color: #fff;
+}
+.fp-chip-card.fp-chip-card-alt .fp-chip-go {
+    color: #4a5662;
+}
+.fp-chip-card.fp-chip-card-alt:hover .fp-chip-go {
+    color: #46d39a;
+}
+
+/* WS4 build consent gate (Proceed/Skip): sits inline in a question-row,
+   side by side, so it stays .fp-chip-card's compact pill-shaped sibling
+   rather than that pattern's full-width stacked card layout. Reuses the
+   same cockpit accent/muted color language so both read as one family. */
+.fp-consent-chip {
+    display: inline-flex;
+    align-items: center;
+    padding: 6px 14px;
+    border-radius: 999px;
+    font-size: 12.5px;
+    font-weight: 600;
+    cursor: pointer;
+    white-space: nowrap;
+    background: #141a21;
+    border: 1px solid #2c3742;
+    color: #dfe6ee;
+    transition: border-color .12s ease, box-shadow .12s ease;
+}
+
+.fp-consent-chip:hover:not(:disabled) {
+    border-color: #46d39a;
+    box-shadow: 0 0 0 1px #46d39a, 0 0 10px -2px #46d39a;
+}
+
+.fp-consent-chip:active:not(:disabled) {
+    transform: translateY(1px);
+}
+
+.fp-consent-chip:disabled {
+    cursor: default;
+    opacity: 0.75;
+}
+
+.fp-consent-chip-primary {
+    border-color: rgba(70, 211, 154, 0.35);
+    color: #46d39a;
+}
+
+.fp-consent-chip-alt {
+    background: transparent;
+    border-color: #2a333c;
+    color: #8a96a3;
+}
+
 .fp-chip-hint {
     margin-top: 4px;
     font-size: 12px;
@@ -596,6 +693,63 @@
     justify-content: flex-end;
 }
 
+/* W4 todo spine. Single-item plans render as a status line (.fp-todo-status);
+   multi-item plans render as a card (.fp-todo-card) with a checklist inside. */
+.fp-todo-status {
+    display: inline-block;
+    font-size: 12px;
+    padding: 3px 10px;
+    margin: 0 0 4px 0;
+    border-radius: 12px;
+    max-width: 100%;
+    overflow: hidden;
+    text-overflow: ellipsis;
+    white-space: nowrap;
+}
+.fp-todo-pending, .fp-todo-active {
+    background: var(--red-ui-secondary-background, #eee);
+    color: var(--red-ui-secondary-text-color, #888);
+}
+.fp-todo-done {
+    background: rgba(34, 139, 34, 0.12);
+    color: #1a7a1a;
+}
+.fp-todo-failed {
+    background: rgba(200, 40, 40, 0.10);
+    color: #b22222;
+}
+.fp-todo-card {
+    padding: 6px 10px;
+    margin: 0 0 4px 0;
+    background: var(--red-ui-secondary-background, #eee);
+    border-radius: 6px;
+    font-size: 13px;
+}
+.fp-todo-list {
+    list-style: none;
+    padding: 0;
+    margin: 0;
+}
+.fp-todo-item {
+    padding: 1px 0;
+    line-height: 1.5;
+}
+.fp-todo-item-pending { color: var(--red-ui-secondary-text-color, #999); }
+.fp-todo-item-active  { font-weight: 600; color: var(--red-ui-text-color, #333); }
+.fp-todo-item-done    { color: #1a7a1a; }
+.fp-todo-item-failed  { color: #b22222; }
+
+/* P10-D2: honest-interruption notice appended inside a "todo" record whose
+   last run event isn't "done" (see rerenderTodoRecord, modes.js). */
+.fp-todo-interrupted {
+    margin-top: 6px;
+    padding: 4px 8px;
+    border-radius: 4px;
+    font-size: 12px;
+    background: rgba(200, 140, 0, 0.10);
+    color: #8a5a00;
+}
+
 .fp-json-toolbar {
     display: flex;
     justify-content: flex-end;