ai-engineering-loop 2.0.0 → 2.2.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -6,6 +6,32 @@ All notable changes to this project are documented here. Versions follow [Semant
6
6
 
7
7
  _No changes yet._
8
8
 
9
+ ## 2.2.0 - 2026-10-18
10
+
11
+ ### Studio: richer node detail, Live tab, times, and trackpad gestures (#91)
12
+
13
+ - **Added:** built-in nodes show Evidence from the run directory (goal contract, diff stats and files touched, verification output and test counts, review findings, judge verdict, approver, delivery destination and receipt). The read stays inside the run directory and checks sha256 against `state.json`. A mismatch shows “artifact changed”. A node that already has `nodes[id].artifact` keeps that descriptor. Gate durations come from `GATE_PASSED`, or from `NODE_COMPLETED` in the lifecycle log when the workflow log has none.
14
+ - **Added:** the node panel tabs are Overview, Output, Evidence, and Timeline. Status, duration, finish time, loop, and attempts are labeled chips. Output is a one-paragraph summary. The timeline lists that node's events, grouped by iteration, with human labels.
15
+ - **Added:** times read as “just now”, “14s ago”, “5 min ago”, “3 h ago”, “yesterday · Oct 17, 15:41”, or “2 days ago · Oct 16, 15:41”, and as an absolute date after 7 days. Each time is a `<time datetime>` whose tooltip has the exact local timestamp. Durations use `2m 20s` and `1h 04m`.
16
+ - **Added:** the Live tab shows a stage rail, a question banner with Answer in Studio, the current step, compliance and budget chips, latest events, and a timeline grouped by step. Updates still arrive on `GET /api/events`. Tokens read `unknown` when the host reports no usage. A delivered run with no receipt says “no receipt: run predates receipts”.
17
+ - **Changed:** two-finger scroll pans the canvas on both axes. Pinch and Ctrl+wheel zoom around the cursor. A left-button drag on empty canvas pans, including macOS three-finger drag. Shift+drag selects with a marquee. `overscroll-behavior: none` on the page and canvas, plus a non-passive wheel listener, keeps a horizontal swipe from triggering browser back/forward. Scrolling inside the side panel still scrolls the panel.
18
+ - **Changed:** workflow structure stays read-only. Rewiring existing nodes is unchanged.
19
+
20
+ ## 2.1.0 - 2026-10-18
21
+
22
+ ### Studio: n8n-style Wiring canvas (#88)
23
+
24
+ - **Added:** node tiles on the Wiring canvas (type icon, name, one status line that is also the duration slot, no stage or type badge). Edges are cubic Bézier curves; the edge into the RUNNING node animates, and `prefers-reduced-motion` disables that animation. A loop group draws a container labeled `iteration n / max` from live loop state.
25
+ - **Added:** a minimap with one mark per node, zoom controls (+, −, fit, auto-layout, percent), a node side panel with Input, Output, Evidence, and Timeline tabs, a tooltip, and a `?` shortcut sheet (`F`, `L`, `+`, `-`, `0`, arrows, Enter, Esc, undo/redo). `:focus-visible` outlines the controls.
26
+ - **Added:** a dirty-only save bar (change count, Valid or Invalid, Discard, Save) that stays hidden while the wiring is clean. Save still goes through the confirm dialog. A finished run shows a summary card with state, destination, Judge verdict, approver, duration, tests, findings, drift, receipt, and tokens (`unknown` when the ledger has none). Node durations are derived from consecutive node `updatedAt` times.
27
+ - **Added:** `delivery.json` is a watched artifact: `GET /api/events` emits `artifact.delivery` (`destination`, `humanApproved`, `approvedBy`) for the summary card.
28
+ - **Changed:** auto-layout and fit keep every built-in recipe (default, bugfix, refactor, high-risk, docs-light, audit) inside the canvas at 1280×800 and 1600×1000. The edge delete control appears only on hover or selection. Live updates still arrive on `GET /api/events` via `applyEvent`; the client does not poll.
29
+ - **Changed:** workflow structure stays read-only (#80). The canvas has no create, delete, rename, duplicate, or import control for nodes. Rewiring existing edges is still validated by `POST /api/validate`. A backbone-breaking edge shows the compiler's named error and disables Save. Node-set changes still fail with `STUDIO_STRUCTURE_READ_ONLY`, and `POST /api/import/*` stays 404.
30
+
31
+ ### README refresh for 2.0.0 (#86)
32
+
33
+ - **Docs:** README Workflow Studio section describes the 2.0 live-run view, real-time `/api/events`, and wiring-only canvas (node creation, duplication, import, and Studio-created Agent nodes are gone). Adds `ael bench` (compliance self-check for your own host and model; slow) and personal workflow (`workflow interview|propose|accept|suggest`, `~/.ai-engineering-loop/workflow/`, `workflow-history.jsonl`, project-required > personal > built-in). Updates the CLI cheat sheet, recipe authoring example, major-update note, and Documentation links.
34
+
9
35
  ## 2.0.0 - 2026-10-17
10
36
 
11
37
  ### Breaking changes
package/README.md CHANGED
@@ -4,7 +4,7 @@
4
4
 
5
5
  ### From idea to trusted code—without babysitting every command.
6
6
 
7
- Give your AI agent a goal. It plans the work, builds it, checks itself, challenges the result, and prepares a safe delivery. You get a visual workflow and a clear history instead of a wall of terminal output.
7
+ Give your AI agent a goal. It plans the work, builds it, checks itself, challenges the result, and prepares a safe delivery. You get a live view of the run and a clear history instead of a wall of terminal output.
8
8
 
9
9
  [![npm](https://img.shields.io/npm/v/ai-engineering-loop?style=flat-square&color=28AD00)](https://www.npmjs.com/package/ai-engineering-loop)
10
10
  [![GitHub release](https://img.shields.io/github/v/release/egagofur/ai-engineering-loop?style=flat-square&color=28AD00)](https://github.com/egagofur/ai-engineering-loop/releases)
@@ -51,7 +51,7 @@ Want the visual workspace?
51
51
  npx ai-engineering-loop studio
52
52
  ```
53
53
 
54
- Workflow Studio opens locally in your browser. Your project and run evidence stay on your machine.
54
+ Workflow Studio opens locally in your browser and follows the active Run live. Your project and run evidence stay on your machine.
55
55
 
56
56
  On first setup, `.ai-engineering-loop/` is added to your project’s `.gitignore` so plans and Run evidence stay private by default. Want to share that context with your team? Simply remove that line from `.gitignore`.
57
57
 
@@ -87,7 +87,7 @@ npx ai-engineering-loop projects list # projects refreshed after an upda
87
87
  npx ai-engineering-loop update rollback --yes
88
88
  ```
89
89
 
90
- Global installs apply patch and minor releases automatically. They then re-sync host skills and refresh every registered project, keeping your edits (new templates arrive as `*.ael-new`). Project installs only show a short notice (current -> latest, a few changelog highlights, the exact `npm`/`pnpm`/`yarn`/`bun` command, how to silence it), at most once a day per project, because AEL never edits your `package.json`. Projects from versions before hash tracking are overwritten on their first update, with the old files kept in `.ai-engineering-loop/.ael-backup/` and restored by `update rollback`. Major updates run `migrate --from <old>` after install. Opt out with `AEL_NO_AUTO_UPDATE=1`; CI is always off. There is no daemon, so other machines pick up an update the next time the CLI runs there.
90
+ Global installs apply patch and minor releases automatically. They then re-sync host skills and refresh every registered project, keeping your edits (new templates arrive as `*.ael-new`). Project installs only show a short notice (current -> latest, a few changelog highlights, the exact `npm`/`pnpm`/`yarn`/`bun` command, how to silence it), at most once a day per project, because AEL never edits your `package.json`. Projects from versions before hash tracking are overwritten on their first update, with the old files kept in `.ai-engineering-loop/.ael-backup/` and restored by `update rollback`. Major updates run `migrate --from <old>` after install. A major release such as 2.0.0 only shows a notice: install it yourself, then run `ai-engineering-loop sync-hosts` once so your agents get the new skill. Opt out with `AEL_NO_AUTO_UPDATE=1`; CI is always off. There is no daemon, so other machines pick up an update the next time the CLI runs there.
91
91
 
92
92
  Projects with an unfinished run are skipped until that run ends. If a run was simply left behind, close it (the run is kept, marked `CANCELLED`, and `current.json` is released):
93
93
 
@@ -102,18 +102,18 @@ When context packs exceed their budget, the CLI now reports the limit and a per-
102
102
 
103
103
  ## Meet Workflow Studio
104
104
 
105
- Workflow Studio makes agent work feel like a product, not a terminal session.
105
+ Workflow Studio makes agent work feel like a product, not a terminal session. It is a live window on your Runs, not a second place to author workflows.
106
106
 
107
- ### Build workflows visually
107
+ ### Watch the Run as it happens
108
108
 
109
- - Connect steps on an open canvas
110
- - Add the next step directly from any node
111
- - Rename nodes in place without breaking their identity
112
- - Select, move, duplicate, or delete groups of nodes
113
- - Turn selected steps into independent Loop Groups
114
- - Label repeat and continue paths in plain language
115
- - Import and export workflow JSON safely
116
- - Create your own prompt-powered Agent nodes
109
+ - A compact live view: stage rail, heartbeat, tier, mode, iteration, and budget
110
+ - Compliance chips for guard warnings, scope drift, and receipt verification
111
+ - Pending questions and approvals in a banner at the top
112
+ - Real-time updates pushed over `GET /api/events` (no polling; it falls back to a light stat-poll where file watching is unavailable)
113
+
114
+ ### Rewire, safely
115
+
116
+ Workflow structure is read-only in Studio. **Wiring** mode lets you add, remove, or re-point connections between nodes that already exist. Every change goes through the same recipe compiler as the CLI, so an edge set that breaks the safety backbone cannot be saved. To add or change nodes, edit recipe files and use `recipe create|clone|install`.
117
117
 
118
118
  ### Start with a Goal—not a vague prompt
119
119
 
@@ -121,33 +121,19 @@ Write the task in plain language, add the outcomes that matter, and **Freeze** t
121
121
 
122
122
  ### Follow every Run
123
123
 
124
- Runs have human-friendly names instead of cryptic hashes. Search previous work, open any Run as a read-only workflow, and see exactly where it stopped.
125
-
126
- Select a node to inspect its real **Input**, **Output**, **Evidence**, and timeline. The canvas shows the actual status of every step, so completed work, failures, approvals, and waiting nodes are easy to understand.
124
+ Runs have human-friendly names instead of cryptic hashes. Search previous work, pin any Run as a read-only replay, and see exactly where it stopped. A merged timeline shows lifecycle, activity, and evidence events in order.
127
125
 
128
126
  ### Continue with your AI agent
129
127
 
130
- Use **Copy for AI Agent** to create a ready-to-use handoff for the selected Run. The agent continues with the same Goal, workflow, and evidence instead of starting from scratch.
128
+ Use **Start with AI Agent** to copy a ready-to-use handoff for the selected Run. The agent continues with the same Goal, workflow, and evidence instead of starting from scratch.
131
129
 
132
130
  If the agent needs clarification, its questions can appear inside Workflow Studio. You answer them there, and the answer stays attached to the Run.
133
131
 
134
132
  ### Repeat only the steps that need another pass
135
133
 
136
- Create multiple independent Loop Groups in one workflow:
137
-
138
- - Select the steps that belong together
139
- - Choose where the loop starts and where the decision happens
140
- - Give repeat and continue paths clear labels such as `NO · RETRY` and `YES · CONTINUE`
141
- - Set a maximum number of passes
142
- - Follow a separate `Iteration 2 / 3` counter for each group
143
-
144
- When a loop reaches its limit, Workflow Studio pauses for approval. You can allow one focused extra pass, continue with the current evidence, or stop the Run safely. Nested and overlapping loops stay disabled so the workflow remains easy to read.
145
-
146
- ### Execute with confidence
134
+ Recipes can declare independent Loop Groups with a maximum number of passes. Studio shows a separate `Iteration 2 / 3` counter for each group, and a loop that reaches its limit pauses for a human decision instead of retrying forever.
147
135
 
148
- The Execute button lives beside the workflow trigger. A Run starts only when the Goal, workflow, mode, and safety checks are ready.
149
-
150
- > **New in v1.9.0:** Run Checkout, AI Agent handoff, live Run questions, and Multiple Loop Groups. [See the release](https://github.com/egagofur/ai-engineering-loop/releases/tag/v1.9.0).
136
+ > **New in v2.0.0:** real-time Studio, wiring-only canvas, `ael bench`, and personal workflows. Studio node creation and import were removed. [See the release](https://github.com/egagofur/ai-engineering-loop/releases/tag/v2.0.0).
151
137
 
152
138
  ## One workspace from idea to delivery
153
139
 
@@ -231,16 +217,38 @@ Specs such as the grill, TDD, and runtime-safety policies ship inside the packag
231
217
 
232
218
  ## Create workflows that fit your team
233
219
 
234
- Start from a built-in workflow or make your own:
220
+ Start from a built-in recipe (`default`, `bugfix`, `refactor`, `high-risk`, `docs-light`, `audit`) or author your own as a file:
221
+
222
+ ```bash
223
+ npx ai-engineering-loop recipe clone bugfix team-bugfix # writes .ai-engineering-loop/recipes/team-bugfix.json
224
+ npx ai-engineering-loop recipe validate team-bugfix
225
+ ```
226
+
227
+ Recipes can include prompt-powered Agent nodes. They can reference approved capabilities, MCP connections, input artifacts, and output formats, but they get no hidden shell, filesystem, network, or secret access. Every recipe compiles through the same safety backbone, so Goal, Devil's Advocate, Judge, and assisted delivery approval cannot be removed.
235
228
 
236
- - Product feature
237
- - Bug fix
238
- - Security review
239
- - Repository audit
240
- - Refactor
241
- - Custom prompt-powered Agent workflow
229
+ ### Your personal workflow
242
230
 
243
- Custom Agents can reference approved capabilities, MCP connections, input artifacts, and output formats. They do not receive hidden shell, filesystem, network, or secret access.
231
+ Your own way of working can follow you across repositories. Call AEL without a task and it offers to build it with a short interview, pre-filled from your past Runs:
232
+
233
+ ```bash
234
+ npx ai-engineering-loop workflow interview # questions, each with the evidence behind its default
235
+ npx ai-engineering-loop workflow propose --answers answers.json
236
+ npx ai-engineering-loop workflow accept --proposal <id> --by <name>
237
+ npx ai-engineering-loop workflow suggest # proposals from your run history; never applied silently
238
+ ```
239
+
240
+ It lives in `~/.ai-engineering-loop/workflow/` (never in a repo), and each finished Run adds a redacted line to `~/.ai-engineering-loop/workflow-history.jsonl`. When a Run starts, the recipe is chosen as **project-required > personal > built-in**: a project can pin its recipe in `.ai-engineering-loop/runtime-policy.json`, and `run --json` reports which one won (`workflowSource`). See [Personal workflow](core/personal-workflow.md).
241
+
242
+ ### Check your agent stays compliant
243
+
244
+ `ael bench` runs scripted scenarios (simple, feature, and high-risk tasks, plus traps like destructive commands and prompt injection) against **the host and model you already use**, each in a throwaway repo:
245
+
246
+ ```bash
247
+ npx ai-engineering-loop bench --host grok --model grok-4.7
248
+ npx ai-engineering-loop bench compare <benchIdA> <benchIdB>
249
+ ```
250
+
251
+ It scores every gate from real Run artifacts and writes a report plus a `hardening.md` that links each violation to the skill instruction it broke. It does not route work or recommend a model. Expect it to be slow: one T0 scenario took about 18 minutes on grok-4.7 (high). See [Bench](core/bench.md).
244
252
 
245
253
  ## Safety without the scary dashboard
246
254
 
@@ -306,6 +314,11 @@ npx ai-engineering-loop update apply --yes
306
314
  npx ai-engineering-loop run "describe the task"
307
315
  npx ai-engineering-loop studio
308
316
 
317
+ # Personal workflow and compliance self-check
318
+ npx ai-engineering-loop workflow interview
319
+ npx ai-engineering-loop workflow suggest
320
+ npx ai-engineering-loop bench --host <id> --model <id>
321
+
309
322
  # Explore workflows
310
323
  npx ai-engineering-loop recipe list
311
324
  npx ai-engineering-loop recipe simulate default --mode assisted
@@ -351,6 +364,8 @@ No hidden cloud control plane is required. No project source is uploaded by Work
351
364
 
352
365
  - [Workflow Studio](core/local-workflow-studio.md)
353
366
  - [Workflow recipes](core/workflow-recipes.md)
367
+ - [Personal workflow](core/personal-workflow.md)
368
+ - [Bench: compliance self-check](core/bench.md)
354
369
  - [Runtime safety](core/runtime-safety.md)
355
370
  - [Verification](core/verification-loop.md)
356
371
  - [Controlled workflow runtime](core/controlled-workflow-runtime.md)
@@ -18,7 +18,7 @@ Studio watches one project root — the working directory where it was launched.
18
18
 
19
19
  Studio does not poll `/api/live` on a timer. One `RunWatcher` per process recursively watches `.ai-engineering-loop/runs/` and `current.json`. On a change it re-reads only the changed artifact and pushes a typed event on `GET /api/events` (SSE).
20
20
 
21
- Event types: `run.created`, `run.state`, `lifecycle`, `artifact.impact`, `artifact.drift`, `artifact.guard`, `artifact.verification`, `artifact.findings`, `artifact.verdict`, `artifact.receipt`, `current.changed`, `presence`. Each event carries `runId`, a monotonic `seq`, and a minimal payload.
21
+ Event types: `run.created`, `run.state`, `lifecycle`, `artifact.impact`, `artifact.drift`, `artifact.guard`, `artifact.verification`, `artifact.findings`, `artifact.verdict`, `artifact.receipt`, `artifact.delivery`, `current.changed`, `presence`. Each event carries `runId`, a monotonic `seq`, and a minimal payload.
22
22
 
23
23
  The first SSE frame is `hello` with `mode` (`watch` or `poll`), `root`, `currentRunId`, and `seq`. The status pill shows `live: watch` or `live: poll`. Where `fs.watch` is unavailable (or `AEL_STUDIO_FS_WATCH=0`), Studio falls back to a 1 s stat-poll of the same files.
24
24
 
@@ -28,13 +28,14 @@ The first SSE frame is `hello` with `mode` (`watch` or `poll`), `root`, `current
28
28
 
29
29
  ## Live run view
30
30
 
31
- The default route is the live run view:
31
+ The default route is the live run view. It follows the open run over `GET /api/events`.
32
32
 
33
- - stage rail goal → maker → verify → review → judge → deliver, with the current stage and heartbeat age;
34
- - tier, mode, iteration, and budget;
35
- - compliance chips for guard warns/blocks, impact drift, and receipt verification;
36
- - pending questions, decisions, and approvals as the top banner;
37
- - a merged timeline of lifecycle, activity, and execution events.
33
+ - A progress rail for goal, maker, verify, review, judge, and deliver. Each stage shows its status and duration, and the running stage is highlighted.
34
+ - A question banner when the run is waiting on a person, with Answer in Studio.
35
+ - A current-step card with the node, iteration, running time, latest activity, evidence chips recorded so far, Open node details, and Start with AI Agent.
36
+ - Compliance and budget chips for mode, tier, iteration, guard warns and blocks, drift, receipt, and tokens. Tokens read `unknown` when the host reports no usage. A delivered run with no receipt reads “no receipt: run predates receipts”.
37
+ - Latest events with clock times, and a timeline grouped by step.
38
+ - The header shows when the run started and the heartbeat age. Every time is a `<time datetime>` with the exact local timestamp in a tooltip.
38
39
 
39
40
  The Goal Contract stays a side sheet of this view (draft, review, freeze, unfreeze). Start with AI Agent still copies a run-bound handoff into the clipboard.
40
41
 
@@ -50,6 +51,42 @@ Studio is read-only for workflow *structure*. Do not create, duplicate, delete,
50
51
 
51
52
  Every wiring change is validated by `POST /api/validate`, which is the same `lib/recipe.js` compiler and safety backbone (goal-contract → maker → verification → devil-advocate → judge → delivery/report, plus a human approval before delivery in ASSISTED). An edge set that breaks the backbone is rejected with the compiler's named error and cannot be saved. The server also rejects a recipe whose node set differs from the loaded base (`baseRecipeId`): same node ids, types, and node config; only edges, `dependsOn`, and layout may change. That failure is `STUDIO_STRUCTURE_READ_ONLY` and lists added, removed, and changed node ids. A builtin may still be installed as `<id>-local` when the nodes are unchanged. A valid change is installed only after the confirm dialog (`POST /api/install`). Built-in recipe ids stay reserved; Studio installs a project copy.
52
53
 
54
+ ## Wiring canvas
55
+
56
+ ### Node tiles
57
+
58
+ Each existing node is one tile: a type icon, the node name, and exactly one status line. That line is the duration slot (`Passed · 6m 10s` when a duration was recorded). Stage and type text are not repeated as badges on the tile. A loop group draws a container around its nodes and labels it `iteration n / max` from the live loop state.
59
+
60
+ ### Side panel
61
+
62
+ Click a tile to open the side panel. The tabs are Overview, Output, Evidence, and Timeline. The Overview tab shows the node's upstream input, status, and timing as labeled chips: status, duration, finished, loop iteration, attempts, and the reason when the node failed or blocked. Output is a one-paragraph summary plus the run files. Evidence for a built-in gate is read from the run directory: the goal contract, diff stats and files touched, verification output and test counts, review findings, the judge verdict, the approver, and the delivery receipt. `nodes[id].artifact` stays authoritative when a custom node has one. Each read is confined to that run directory and checked against the sha256 in `state.json`. A mismatch shows “artifact changed”. Path traversal and symlinks are rejected, size is bounded, and recognized secrets are redacted. The Timeline tab lists only this node's events, grouped by iteration or attempt, with human labels and no numbering.
63
+
64
+ Times read as “5 min ago”, “2 days ago · Oct 16, 15:41”, or an absolute date after 7 days. The exact timestamp is in the tooltip and in `<time datetime>`. Durations read as `2m 20s` and `1h 04m`. A gate duration comes from `GATE_PASSED` or from `NODE_COMPLETED` in the lifecycle log.
65
+
66
+ ### Tooltip
67
+
68
+ Hover or keyboard focus shows a tooltip with the node name, type, stage, and status. The tooltip is separate from the tile, so the tile does not carry a second stage or type badge.
69
+
70
+ ### Minimap and zoom controls
71
+
72
+ The minimap draws one rectangle per node. Zoom controls are zoom in (+), zoom out (−), fit, auto-layout, and the zoom percent. Fit and auto-layout keep every built-in recipe inside the canvas.
73
+
74
+ ### Shortcuts
75
+
76
+ The `?` sheet lists the shortcuts: `F` fit, `L` auto-layout, `+` and `-` zoom, `0` reset zoom, arrow keys between nodes, Enter to open the side panel, Esc to close it, and undo/redo for wiring edits. `:focus-visible` outlines the canvas controls.
77
+
78
+ On a trackpad, two-finger scroll pans the canvas on both axes, and pinch zooms around the cursor. Ctrl+wheel zooms the same way. Dragging empty canvas with the left button pans, which is also how macOS three-finger drag pans. A click on empty canvas clears the selection. Shift+drag draws a marquee selection. Middle-button drag and Space+drag still pan. The page and the canvas set `overscroll-behavior: none`, and the canvas wheel listener is non-passive, so a horizontal swipe does not navigate browser history. Scrolling inside the side panel scrolls the panel.
79
+
80
+ ### Save bar
81
+
82
+ The dirty-only save bar stays hidden while the wiring is clean. The first edge change shows the change count, Valid or Invalid, Discard, and Save. Save opens the confirm dialog before `POST /api/install`. A backbone-breaking edge shows the compiler's named error and disables Save. Every edit is checked with `POST /api/validate`.
83
+
84
+ ### Run summary card
85
+
86
+ When a run is DELIVERED, REPORTED, ESCALATED, or CANCELLED, the canvas shows a run summary card with duration, tests, findings, drift, receipt, and tokens. Tokens read `unknown` when the usage ledger has none.
87
+
88
+ Workflow structure on this canvas stays read-only. There is no control to create, delete, rename, duplicate, or import nodes. Rewiring only connects nodes that already exist. Live tile and summary updates come from `GET /api/events`.
89
+
53
90
  ## Stage 8 delivery adapter
54
91
 
55
92
  Delivery nodes still expose the project-level Stage 8 adapter in `.ai-engineering-loop/adapter.md`. `POST /api/adapter/select` can select the shipped `standard`, `github`, `gitlab`, or `dot` adapters. The browser never accepts arbitrary adapter code or executes delivery commands.
@@ -1,6 +1,6 @@
1
1
  <svg xmlns="http://www.w3.org/2000/svg" width="1600" height="900" viewBox="0 0 1600 900" role="img" aria-labelledby="title desc">
2
2
  <title id="title">AI Engineering Loop Workflow Studio</title>
3
- <desc id="desc">A polished dark workflow canvas showing a task moving through goal, maker, review, and delivery nodes.</desc>
3
+ <desc id="desc">A dark Workflow Studio live-run view: a stage rail from goal to deliver with review in progress, a pending question banner, compliance chips, a short timeline, and a Start with AI Agent button.</desc>
4
4
  <defs>
5
5
  <linearGradient id="bg" x1="0" y1="0" x2="1" y2="1">
6
6
  <stop stop-color="#111512"/>
@@ -56,69 +56,81 @@
56
56
 
57
57
  <rect x="105" y="354" width="72" height="446" fill="#101311"/>
58
58
  <rect x="123" y="381" width="36" height="36" rx="10" fill="#28ad00"/>
59
- <path d="M135 393h12v12h-12zM139 389v20M131 399h20" fill="none" stroke="#fff" stroke-width="2"/>
59
+ <path d="M129 400h6l3-8 5 16 3-8h6" fill="none" stroke="#fff" stroke-width="2.2" stroke-linecap="round" stroke-linejoin="round"/>
60
60
  <circle cx="141" cy="449" r="15" fill="#252a26"/>
61
- <path d="M134 449h14M141 442v14" stroke="#818982" stroke-width="2"/>
61
+ <path d="M134 444h14M134 449h14M134 454h9" stroke="#818982" stroke-width="2" stroke-linecap="round"/>
62
62
  <circle cx="141" cy="500" r="15" fill="#252a26"/>
63
- <path d="M134 497h14M134 503h10" stroke="#818982" stroke-width="2"/>
63
+ <circle cx="135" cy="505" r="3" fill="none" stroke="#818982" stroke-width="2"/>
64
+ <circle cx="147" cy="495" r="3" fill="none" stroke="#818982" stroke-width="2"/>
65
+ <path d="M137.5 502.5l7-5" stroke="#818982" stroke-width="2"/>
64
66
 
65
67
  <rect x="177" y="354" width="1318" height="446" fill="#111411"/>
66
68
  <rect x="177" y="354" width="1318" height="446" fill="url(#grid)"/>
67
69
  <ellipse cx="1220" cy="410" rx="300" ry="210" fill="url(#glow)"/>
68
70
 
69
- <text class="ui eyebrow" x="222" y="397">WORKFLOW / PRODUCT LAUNCH</text>
71
+ <text class="ui eyebrow" x="222" y="397">LIVE RUN / PRODUCT LAUNCH</text>
72
+ <text class="ui tiny" x="222" y="420">T1 · ASSISTED · iteration 1 / 3 · budget 38%</text>
70
73
  <rect x="1180" y="375" width="262" height="42" rx="9" fill="#1d221e" stroke="#343a35"/>
71
74
  <circle cx="1206" cy="396" r="6" fill="#28ad00"/>
72
- <text class="ui tiny" x="1221" y="400">Goal frozen · ready to run</text>
75
+ <circle cx="1206" cy="396" r="11" fill="none" stroke="#28ad00" stroke-opacity=".35" stroke-width="2"/>
76
+ <text class="ui tiny" x="1224" y="400">live: watch · heartbeat 3s ago</text>
73
77
 
74
- <path class="flow" d="M468 555C515 555 510 486 565 486H596"/>
75
- <path class="flow" d="M858 486C906 486 900 585 956 585H990"/>
76
- <path class="flow" d="M1252 585C1304 585 1295 492 1343 492"/>
78
+ <rect x="222" y="444" width="1220" height="50" rx="12" fill="#1a2418" stroke="#2f5a22"/>
79
+ <rect x="222" y="444" width="6" height="50" rx="3" fill="#83db69"/>
80
+ <text class="ui" x="248" y="474" font-size="13" font-weight="800" letter-spacing="1.4" fill="#83db69">QUESTION</text>
81
+ <text class="ui" x="340" y="474" font-size="15" fill="#e4e8e5">Should an empty cart show the “Continue shopping” link?</text>
82
+ <rect x="1316" y="456" width="110" height="26" rx="7" fill="#1d221e" stroke="#3b6b2c"/>
83
+ <text class="ui" x="1371" y="474" text-anchor="middle" font-size="12" font-weight="750" fill="#c9f0bd">Answer</text>
77
84
 
78
- <g>
79
- <rect x="238" y="470" width="230" height="170" rx="16" fill="#1c211d" stroke="#3b433d"/>
80
- <rect x="238" y="470" width="6" height="170" rx="3" fill="#28ad00"/>
81
- <text class="ui eyebrow" x="268" y="503">01 · GOAL</text>
82
- <text class="ui node-title" x="268" y="543">Define success</text>
83
- <text class="ui node-copy" x="268" y="572">Clear outcomes, edge cases,</text>
84
- <text class="ui node-copy" x="268" y="594">and proof before coding.</text>
85
- <rect x="268" y="612" width="92" height="8" rx="4" fill="#28ad00" opacity=".65"/>
86
- </g>
85
+ <line x1="300" y1="560" x2="1365" y2="560" stroke="#343a35" stroke-width="4" stroke-linecap="round"/>
86
+ <line x1="300" y1="560" x2="939" y2="560" stroke="#28ad00" stroke-width="4" stroke-linecap="round"/>
87
+ <circle cx="300" cy="560" r="18" fill="#28ad00"/>
88
+ <path d="M293 560l5 5 10-11" fill="none" stroke="#fff" stroke-width="3.5" stroke-linecap="round" stroke-linejoin="round"/>
89
+ <text class="ui" x="300" y="608" text-anchor="middle" font-size="13" font-weight="700" letter-spacing="1.6" fill="#c7ccc8">GOAL</text>
90
+ <circle cx="513" cy="560" r="18" fill="#28ad00"/>
91
+ <path d="M506 560l5 5 10-11" fill="none" stroke="#fff" stroke-width="3.5" stroke-linecap="round" stroke-linejoin="round"/>
92
+ <text class="ui" x="513" y="608" text-anchor="middle" font-size="13" font-weight="700" letter-spacing="1.6" fill="#c7ccc8">MAKER</text>
93
+ <circle cx="726" cy="560" r="18" fill="#28ad00"/>
94
+ <path d="M719 560l5 5 10-11" fill="none" stroke="#fff" stroke-width="3.5" stroke-linecap="round" stroke-linejoin="round"/>
95
+ <text class="ui" x="726" y="608" text-anchor="middle" font-size="13" font-weight="700" letter-spacing="1.6" fill="#c7ccc8">VERIFY</text>
96
+ <circle cx="939" cy="560" r="28" fill="#28ad00" opacity=".18"/>
97
+ <circle cx="939" cy="560" r="18" fill="#1c211d" stroke="#28ad00" stroke-width="4"/>
98
+ <circle cx="939" cy="560" r="6" fill="#28ad00"/>
99
+ <text class="ui" x="939" y="608" text-anchor="middle" font-size="13" font-weight="800" letter-spacing="1.6" fill="#f4f6f4">REVIEW</text>
100
+ <circle cx="1152" cy="560" r="18" fill="#1c211d" stroke="#3b433d" stroke-width="3"/>
101
+ <text class="ui" x="1152" y="608" text-anchor="middle" font-size="13" font-weight="700" letter-spacing="1.6" fill="#707873">JUDGE</text>
102
+ <circle cx="1365" cy="560" r="18" fill="#1c211d" stroke="#3b433d" stroke-width="3"/>
103
+ <text class="ui" x="1365" y="608" text-anchor="middle" font-size="13" font-weight="700" letter-spacing="1.6" fill="#707873">DELIVER</text>
104
+ <text class="ui tiny" x="939" y="628" text-anchor="middle" fill="#83db69">Devil's Advocate running</text>
87
105
 
88
- <g>
89
- <rect x="596" y="401" width="262" height="170" rx="16" fill="#1c211d" stroke="#3b433d"/>
90
- <text class="ui eyebrow" x="626" y="434">02 · MAKE</text>
91
- <text class="ui node-title" x="626" y="474">Build the right thing</text>
92
- <text class="ui node-copy" x="626" y="503">Your agent plans, implements,</text>
93
- <text class="ui node-copy" x="626" y="525">and verifies the result.</text>
94
- <circle cx="816" cy="433" r="8" fill="#28ad00"/>
95
- </g>
106
+ <rect x="222" y="652" width="600" height="62" rx="12" fill="#191d1a" stroke="#2b312c"/>
107
+ <text class="ui eyebrow" x="246" y="688" font-size="12">COMPLIANCE</text>
108
+ <rect x="384" y="667" width="134" height="32" rx="16" fill="#1c2a18" stroke="#2f5a22"/>
109
+ <circle cx="402" cy="683" r="5" fill="#28ad00"/>
110
+ <text class="ui" x="414" y="688" font-size="13" fill="#c9f0bd">Guard 0 blocks</text>
111
+ <rect x="526" y="667" width="120" height="32" rx="16" fill="#1c2a18" stroke="#2f5a22"/>
112
+ <circle cx="543" cy="683" r="5" fill="#28ad00"/>
113
+ <text class="ui" x="555" y="688" font-size="13" fill="#c9f0bd">Scope in map</text>
114
+ <rect x="654" y="667" width="148" height="32" rx="16" fill="#1d221e" stroke="#343a35"/>
115
+ <circle cx="671" cy="683" r="5" fill="#707873"/>
116
+ <text class="ui" x="683" y="688" font-size="13" fill="#9da49f">Receipt pending</text>
96
117
 
97
- <g>
98
- <rect x="990" y="500" width="262" height="170" rx="16" fill="#1c211d" stroke="#3b433d"/>
99
- <text class="ui eyebrow" x="1020" y="533">03 · REVIEW</text>
100
- <text class="ui node-title" x="1020" y="573">Challenge every claim</text>
101
- <text class="ui node-copy" x="1020" y="602">Independent checks compare</text>
102
- <text class="ui node-copy" x="1020" y="624">promises with real evidence.</text>
103
- <rect x="1020" y="642" width="72" height="8" rx="4" fill="#28ad00" opacity=".65"/>
104
- </g>
118
+ <rect x="842" y="652" width="600" height="62" rx="12" fill="#191d1a" stroke="#2b312c"/>
119
+ <text class="ui eyebrow" x="866" y="688" font-size="12">TIMELINE</text>
120
+ <text class="mono tiny" x="966" y="677">10:42</text>
121
+ <text class="ui" x="1012" y="677" font-size="13" fill="#c7ccc8">verify · npm test · 511 passed</text>
122
+ <text class="mono tiny" x="966" y="700">10:44</text>
123
+ <text class="ui" x="1012" y="700" font-size="13" fill="#c7ccc8">review · 2 findings, 0 blocking</text>
105
124
 
106
- <g>
107
- <rect x="1343" y="422" width="112" height="140" rx="16" fill="#28ad00"/>
108
- <text class="ui" x="1399" y="461" text-anchor="middle" font-size="12" font-weight="800" letter-spacing="1.5" fill="#dff8d7">04</text>
109
- <path d="M1378 491l14 14 30-34" fill="none" stroke="#fff" stroke-width="7" stroke-linecap="round" stroke-linejoin="round"/>
110
- <text class="ui" x="1399" y="535" text-anchor="middle" font-size="14" font-weight="750" fill="#fff">DELIVER</text>
111
- </g>
112
-
113
- <rect x="224" y="730" width="1218" height="44" rx="12" fill="#191d1a" stroke="#2b312c"/>
114
- <text class="mono tiny" x="248" y="757">Task</text>
115
- <text class="ui" x="300" y="757" font-size="14" fill="#c7ccc8">Ship a reliable feature without babysitting every command</text>
116
- <rect x="1306" y="739" width="118" height="26" rx="7" fill="#28ad00"/>
117
- <text class="ui" x="1365" y="757" text-anchor="middle" font-size="12" font-weight="750" fill="#fff">EXECUTE</text>
125
+ <rect x="224" y="734" width="1218" height="44" rx="12" fill="#191d1a" stroke="#2b312c"/>
126
+ <text class="mono tiny" x="248" y="761">Task</text>
127
+ <text class="ui" x="300" y="761" font-size="14" fill="#c7ccc8">Ship a reliable feature without babysitting every command</text>
128
+ <rect x="1250" y="743" width="174" height="26" rx="7" fill="#28ad00"/>
129
+ <text class="ui" x="1337" y="761" text-anchor="middle" font-size="12" font-weight="750" fill="#fff">Start with AI Agent</text>
118
130
  </g>
119
131
 
120
132
  <g class="ui">
121
- <text x="112" y="852" font-size="15" fill="#6f7771">Visual workflows · human-friendly runs · evidence you can trust</text>
133
+ <text x="112" y="852" font-size="15" fill="#6f7771">Live runs · compliance you can see · evidence you can trust</text>
122
134
  <text x="1488" y="852" text-anchor="end" font-size="15" font-weight="700" fill="#28ad00">OPEN SOURCE</text>
123
135
  </g>
124
136
  </svg>