@imagile-dev/toil-ui 0.1.0-beta.3 → 0.1.0-beta.4
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/README.md +276 -195
- package/dist/app/assets/{Graph-CFuW4gKS.js → Graph-8jUpxY5T.js} +2 -2
- package/dist/app/assets/{PowershellMan-BMVCrjjz.js → PowershellMan-D_GfPDDt.js} +1 -1
- package/dist/app/assets/index-BJY6F1XS.js +504 -0
- package/dist/app/assets/index-BWejFOaT.css +1 -0
- package/dist/app/assets/theme--CDBXtCV.css +1 -0
- package/dist/app/assets/theme-Bd9rCh2b.css +1 -0
- package/dist/app/assets/theme-C7Mxc3ZD.css +1 -0
- package/dist/app/assets/{theme-mWrwQOT7.css → theme-elgfmZWW.css} +1 -1
- package/dist/app/assets/theme-hbyLw_lf.css +1 -0
- package/dist/app/index.html +2 -2
- package/dist/server/toil-ui.js +1995 -1418
- package/package.json +2 -2
- package/dist/app/assets/index-BcEh-0qQ.css +0 -1
- package/dist/app/assets/index-D9EQ864e.js +0 -74
- package/dist/app/assets/theme-BJbhJDgp.css +0 -1
- package/dist/app/assets/theme-CC5mVuDH.css +0 -1
- package/dist/app/assets/theme-gYmLCU0M.css +0 -1
- package/dist/app/assets/theme-paULUSRX.css +0 -1
package/README.md
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
# Toil UI
|
|
2
2
|
|
|
3
3
|
A local web app for seeing work across Jira, Azure DevOps and GitHub, read through
|
|
4
|
-
[toil](https://www.npmjs.com/package/@imagile-dev/toil): one dashboard per
|
|
4
|
+
[toil](https://www.npmjs.com/package/@imagile-dev/toil): one dashboard per Toil project,
|
|
5
5
|
with its stories, pull requests, builds and deployments side by side, a pull
|
|
6
6
|
requests view for what is waiting on review, and a daily worklog of what happened.
|
|
7
7
|
|
|
8
8
|
**Status: beta, local only.** It runs on your own machine, binds loopback, and
|
|
9
|
-
reads through toil with your own connections and
|
|
10
|
-
settings pages, and a
|
|
11
|
-
places it changes anything of toil's, and only toil's own connections and
|
|
9
|
+
reads through toil with your own connections and Toil projects, read-only. Its
|
|
10
|
+
settings pages, and a Toil project's name and description edited in place, are the only
|
|
11
|
+
places it changes anything of toil's, and only toil's own connections and Toil projects. The worklog keeps its pages and the notes you add in its own
|
|
12
12
|
data folder, and writes anywhere else only when you export a page. Its Feedback dialog sends a report to Imagile through
|
|
13
13
|
`toil feedback`, only when you approve it. It is a separate package from toil: installing toil
|
|
14
14
|
never installs it. toil's beta commitment to stable commands, flags and output doesn't
|
|
@@ -36,7 +36,7 @@ It prints one URL and keeps running until you press Ctrl+C:
|
|
|
36
36
|
|
|
37
37
|
```text
|
|
38
38
|
Toil UI <version> (beta, local only), on toil <version>
|
|
39
|
-
Syncing through toil, read-only, with your own connections and
|
|
39
|
+
Syncing through toil, read-only, with your own connections and Toil projects. Only its Settings pages, and a Toil project's name and description edited in place, change anything: toil's own connections and toilProjects, when you ask. The worklog keeps its pages and your notes in its data folder, and writes anywhere else only when you export one. Feedback goes to Imagile only when you send it.
|
|
40
40
|
The cache is /home/you/.toil-ui/cache/toil-ui.db, kept 30 days after each sync; toil-ui wipe deletes it.
|
|
41
41
|
|
|
42
42
|
Open this URL in your browser. It carries this launch's token, so keep it to yourself:
|
|
@@ -56,37 +56,48 @@ cookie to every port on `127.0.0.1`, so any other local server would receive the
|
|
|
56
56
|
session. `sessionStorage` belongs to this exact origin (port included) and is never
|
|
57
57
|
sent anywhere. `localStorage` would outlive the tab, so it isn't used.
|
|
58
58
|
|
|
59
|
-
`toil-ui --sample` serves the sample
|
|
59
|
+
`toil-ui --sample` serves the sample Toil projects and the scripted agent that
|
|
60
60
|
[the site's Toil UI page](https://imagile-dev.github.io/toil/ui/) runs on, with
|
|
61
61
|
nothing connected to a tracker.
|
|
62
62
|
|
|
63
63
|
## First run
|
|
64
64
|
|
|
65
|
-
Until you have a
|
|
65
|
+
Until you have a Toil project and it has synced once, the Toil projects view shows a short
|
|
66
66
|
checklist in place of the empty page:
|
|
67
67
|
|
|
68
68
|
1. **Connect a tracker** opens Settings, Connections at the add form. If an agent following the toil skill's `setup.md` already created the connection, you only enter its secret here, on its row.
|
|
69
|
-
2. **Create a
|
|
69
|
+
2. **Create a Toil project** opens Settings, Toil projects at Create a Toil project.
|
|
70
70
|
3. **Sync** starts the first sync, as the Sync panel's Sync now does.
|
|
71
71
|
|
|
72
72
|
Each step ticks itself off from toil's own state: `toil connections list` for a
|
|
73
73
|
connection to Jira, Azure DevOps or GitHub (one made by env vars counts), `toil
|
|
74
|
-
|
|
74
|
+
projects list` for a Toil project, and the Sync panel for the sync. Beside the steps,
|
|
75
75
|
**Try the sample data** shows `npx @imagile-dev/toil-ui --sample` with a Copy
|
|
76
|
-
button, to restart Toil UI on the sample
|
|
76
|
+
button, to restart Toil UI on the sample Toil projects. The checklist goes once the first
|
|
77
77
|
sync finishes. Each step's (i) explains it and links to these docs.
|
|
78
78
|
|
|
79
79
|
## What it reads
|
|
80
80
|
|
|
81
|
-
Each dashboard reads its
|
|
82
|
-
Jira project, the Azure DevOps project and pipelines, the GitHub
|
|
83
|
-
project board, and the OpenShift namespace of each environment. A GitHub
|
|
84
|
-
|
|
85
|
-
|
|
86
|
-
|
|
81
|
+
Each dashboard reads its Toil project the way `toil projects show` describes it: the
|
|
82
|
+
Jira project, the Azure DevOps project and pipelines, the GitHub repositories and
|
|
83
|
+
project board, and the OpenShift namespace of each environment. A Toil project's GitHub
|
|
84
|
+
repositories are the `github` section's `owner` and `repo` (one value, or one per repo
|
|
85
|
+
name), plus its `repos` remotes on github.com or on the host of a repository the section names, and each is read on its own
|
|
86
|
+
with `owner` and `repo` passed explicitly (`src/github-repos.ts`); a repository that
|
|
87
|
+
couldn't be read gets one muted line where its gap shows: "pull requests not read" on the Pull requests view (linking to the failed call), "work items not read" on the Board tab. The same goes for the other
|
|
88
|
+
areas' per-repo values: a Jira `project`, a Bitbucket `project`, `workspace` or `repo`,
|
|
89
|
+
and an Azure DevOps `collection`, `project` or `repo` given as an object of repo name to
|
|
90
|
+
value are each read once per target (one Jira search per project, work items per Azure
|
|
91
|
+
DevOps project, pull requests per repo) with those inputs passed explicitly
|
|
92
|
+
(`src/targets.ts`; a pipeline is tried in each Azure DevOps project in turn, and a repo name the section leaves half-named is shown as not read), because Toil UI has no repo folder for toil to pick one from; a section
|
|
93
|
+
with plain values sends the calls it always did. Only GitHub's remotes are read for more
|
|
94
|
+
targets. A GitHub section with no `projects`
|
|
95
|
+
reads each repository's own issues, open and closed, as its work items, with the
|
|
96
|
+
history of every open one and the 50 most recently closed in each. Pull
|
|
97
|
+
requests are read from every code host the Toil project names (GitHub, Bitbucket Data
|
|
87
98
|
Center and Cloud, Azure DevOps Server and Services) with `pr list`, then `pr get
|
|
88
99
|
--include reviews,checks,commits,files` for every open one and the 20 newest closed or
|
|
89
|
-
merged ones, and mapped to OWI 0.2 changes; a GitHub repository's CODEOWNERS is read
|
|
100
|
+
merged ones (of each GitHub repository), and mapped to OWI 0.2 changes; a GitHub repository's CODEOWNERS is read
|
|
90
101
|
with `github repo file` when an open pull request has no reviewer. Every call runs
|
|
91
102
|
in-process through `@imagile-dev/toil/api`, read-only, and toil's audit log
|
|
92
103
|
records it with `via: "api"` and `caller: "toil-ui"`. A call that fails shows on
|
|
@@ -97,73 +108,83 @@ the read window: the retention period, but at least 6 weeks (42 days), so the
|
|
|
97
108
|
longest view stays whole. A closed item older than that isn't read, and so isn't
|
|
98
109
|
in any view.
|
|
99
110
|
|
|
100
|
-
##
|
|
111
|
+
## Plain words
|
|
101
112
|
|
|
102
|
-
|
|
113
|
+
Screens read in everyday words by default: a pull request is a **code review**, an agent an **AI helper**, with toil's own term in the hover. Facts, numbers, commands and errors read the same either way. **Settings, General** (and the command palette) turn it off, kept in this browser; `?words=standard` in the address does too. Themed wording and themed agent voice are off until chosen.
|
|
114
|
+
|
|
115
|
+
The tabs read **Summary**, **Board**, **What blocks what** (Dependencies), **Code reviews**, **People**, **What happened** (Worklog) and **Past briefings** (Analyses); Analyze is **Get a briefing**. A panel's source reads "From Jira and GitHub", with the commands one step in. **Settings → General** has a **Plain words** switch with an (i), and the command palette has **Use plain words** / **Use toil's own terms**. Quoted titles, the tracker's words and a briefing's own sentences are never reworded.
|
|
116
|
+
|
|
117
|
+
## The Toil projects view
|
|
118
|
+
|
|
119
|
+
Toil UI opens on its Toil projects, one card each, read from the same data the dashboards
|
|
103
120
|
read, so it asks no tracker anything: in the local app that is the cache a sync
|
|
104
|
-
filled, plus `
|
|
121
|
+
filled, plus `projects list` and `projects show` for the Toil projects themselves.
|
|
105
122
|
|
|
106
123
|
- **What each card shows:** the current sprint with its working days left, points
|
|
107
124
|
done and a small burn-up; how many items are in progress and blocked, and how many
|
|
108
125
|
pull requests wait on a review; the people with work in it, as avatars; its sources
|
|
109
126
|
as badges; when it was last synced; and how many work items and pull requests
|
|
110
127
|
changed since this browser last opened its dashboard.
|
|
111
|
-
- **Which card leads:** the
|
|
128
|
+
- **Which card leads:** the Toil project opened most recently in this browser, or, before
|
|
112
129
|
any was opened, the one with the most in motion (in progress, blocked and waiting on
|
|
113
130
|
review together), takes two columns where they fit. On a wide screen the grid goes
|
|
114
131
|
wider than the dashboards and fits more cards across; at 375px it is one column.
|
|
115
132
|
- **What this browser remembers:** when each dashboard was last opened, in
|
|
116
133
|
`localStorage`, never sent anywhere. Without it (a private window) the cards just
|
|
117
134
|
say nothing about "since".
|
|
118
|
-
- **A
|
|
119
|
-
sources as badges, and the sprint's day (The
|
|
135
|
+
- **A Toil project's top row** uses the same parts: the description beside the name, the
|
|
136
|
+
sources as badges, and the sprint's day (The Toil project page, below).
|
|
120
137
|
|
|
121
|
-
### Renaming and describing a
|
|
138
|
+
### Renaming and describing a Toil project
|
|
122
139
|
|
|
123
140
|
The name and the description have a pencil beside them, on the card and in the
|
|
124
|
-
|
|
125
|
-
`toil
|
|
141
|
+
Toil project's top row. It opens an inline field: Enter saves through
|
|
142
|
+
`toil projects describe <toilProject> --name <new name>` or `--description <text>`, run
|
|
126
143
|
on the settings pages' path (below), so toil's own checks apply and the change is in
|
|
127
144
|
its audit log as `caller: "toil-ui"`. Escape cancels, and a refusal shows toil's
|
|
128
|
-
message and fix under the field. Only a
|
|
129
|
-
described; a checked-in one is refused with the `toil
|
|
145
|
+
message and fix under the field. Only a Toil project in your own folder can be renamed or
|
|
146
|
+
described; a checked-in one is refused with the `toil projects import` fix.
|
|
130
147
|
|
|
131
|
-
A rename keeps everything Toil UI holds for the
|
|
148
|
+
A rename keeps everything Toil UI holds for the Toil project: the server moves its cache
|
|
132
149
|
(responses, items, history, sources, sync runs, notes and agent runs), its assignments
|
|
133
150
|
and its worklog pages to the new name once toil has renamed the file, so its
|
|
134
151
|
dashboard opens straight away with no new sync. A rename waits, refused with a
|
|
135
|
-
message, while a sync or an assignment of that
|
|
152
|
+
message, while a sync or an assignment of that Toil project is running, since both read it
|
|
136
153
|
by name as they go. What can't be moved is said beside toil's result, and the next
|
|
137
|
-
sync reads the
|
|
154
|
+
sync reads the Toil project afresh under its new name.
|
|
138
155
|
|
|
139
|
-
## The
|
|
156
|
+
## The Toil project page
|
|
140
157
|
|
|
141
|
-
A
|
|
158
|
+
A Toil project opens on its **Overview**, the briefing, in a frame all of its pages share:
|
|
142
159
|
|
|
143
|
-
- **The top row:**
|
|
144
|
-
|
|
145
|
-
line with the whole of it in its tooltip, and its sources; and the Ask bar with
|
|
146
|
-
**
|
|
147
|
-
- **The tabs**, one address each: **Overview** (`#/
|
|
160
|
+
- **The top row:** on the Overview, a **Diagram** link (`#/project/<name>/job/diagram`) draws the Toil project's repositories as a C4 diagram in Mermaid; the steps that need a folder on disk (`util size`, `git log`, `deps scan`) are reported unchecked. Its name, with the pencil that renames it; **Switch**, for another
|
|
161
|
+
Toil project or the Toil projects view; the sprint and its day; its description, cut to one
|
|
162
|
+
line with the whole of it in its tooltip, and its sources (on a phone, a **Details** button beside the name opens them); and the Ask bar, with the
|
|
163
|
+
**AI helpers** button beside it (The agent, below).
|
|
164
|
+
- **The tabs**, one address each: **Overview** (`#/project/<name>`), **Board**
|
|
148
165
|
(`/board`, with how many of its cards need attention), **Dependencies**
|
|
149
166
|
(`/dependencies`), **Pull requests** (`/prs`), **People** (`/people`), **Worklog**
|
|
150
167
|
(`/worklog`) and **Analyses** (`/analyses`). The Command Center (`/command-center`)
|
|
151
168
|
is the Board tab's **Readiness** view, the stand-up is the Worklog's, and an item's
|
|
152
169
|
page sits under Board, so every address from before the tabs still opens. On a
|
|
153
|
-
narrow screen the tabs scroll sideways
|
|
154
|
-
|
|
155
|
-
|
|
170
|
+
narrow screen the tabs scroll sideways; the Agents button sits beside the Ask bar at
|
|
171
|
+
every width (a compact **AI** with its count on a phone, still named "AI helpers"), so the tabs get the whole row.
|
|
172
|
+
- **The Agents panel**, closed by default at every width: one **Agents** button (**AI helpers** in plain words)
|
|
173
|
+
beside the Ask bar, counting what waits on you (the tracker's inbox, proposals to approve and
|
|
174
|
+
questions to answer, and any assignment or proposal that failed, since a closed panel must not hide a failure). From 1200px wide it opens the panel beside every tab, and remembers
|
|
175
|
+
whether you left it open (in `localStorage`); below that it opens a drawer. The palette's
|
|
176
|
+
**Open the Agents panel** does the same. It holds: the tracker's inbox, then the Toil project's assignments
|
|
156
177
|
in three groups (those that need you, those working, and those gone stale), with the
|
|
157
178
|
finished ones under **Earlier**; then **Renew all stale**, the **Limits**, and this
|
|
158
|
-
|
|
179
|
+
Toil project's three newest analyses, each linking to it (Assigning work to an agent, below). Each card's title is its link; how many run at once, for how long, and what the inbox is are in the heading's (i), and the tracker's raw label (`agent:refine`) is in the Start dialog.
|
|
159
180
|
While an item's drawer is open (`?item=<id>`, The briefing, below), it takes the
|
|
160
181
|
sidebar's column from 1200px wide, and covers the view from the right below that.
|
|
161
|
-
- **Analyses** lists every Analyze kept for the
|
|
182
|
+
- **Analyses** lists every Analyze kept for the Toil project, newest first: when it was made,
|
|
162
183
|
the adapter and model that wrote it, its one-line recommendation, and what changed
|
|
163
184
|
since the one before (items newly blocked or blocked by something new, blockers
|
|
164
185
|
cleared, points moved on an item or in the sprint, items newly cited or no longer
|
|
165
186
|
cited, a new recommendation). What changed is worked out by code, from a snapshot of
|
|
166
|
-
the
|
|
187
|
+
the Toil project's items each analysis keeps, never by asking the agent. **Open** reads one
|
|
167
188
|
again at its own address (`/analyses/<id>`), top-down like the chat, with its
|
|
168
189
|
suggested changes and what it used. The server keeps them in the cache (Sync and the
|
|
169
190
|
cache, below), so they outlast a reload and a restart; with `--sample`,
|
|
@@ -171,23 +192,42 @@ A dossier opens on its **Overview**, the briefing, in a frame all of its pages s
|
|
|
171
192
|
|
|
172
193
|
### The briefing
|
|
173
194
|
|
|
174
|
-
A
|
|
175
|
-
further in, and each step one you choose (#494).
|
|
176
|
-
|
|
177
|
-
|
|
178
|
-
|
|
195
|
+
A Toil project opens on its briefing: one column to read top to bottom, each part a step
|
|
196
|
+
further in, and each step one you choose (#494). The tab's top is the answer (#552):
|
|
197
|
+
|
|
198
|
+
- **The status sentence**, the biggest text on the tab: a word with a shape (On
|
|
199
|
+
track, At risk, Behind, No sprint, or No estimates when no story in the sprint has points) and the rest of the sentence ("Behind: 18 of
|
|
200
|
+
21 points are left, and 3 working days remain."), with an (i). That it is an estimate is the first line of the fold; "Too early in the sprint to judge a pace" stays on the page. One rule
|
|
201
|
+
picks the word (`ui/src/model/status.ts`): On track when the points left fit at
|
|
202
|
+
the pace so far, At risk when at least 75% of them would land or a red reason holds
|
|
203
|
+
work up, Behind below 75%; the fold says 75% is the app's own rule of thumb. Today isn't counted in the working days left, as on the
|
|
204
|
+
Toil project cards.
|
|
205
|
+
- **One line on why**, from the briefing on screen, naming items by their titles (quoted as they are, never reworded by plain words), and
|
|
206
|
+
**How was this worked out?** one step in, with the sums and the rule.
|
|
207
|
+
- **Needs attention**, at most three cards, the same for everyone: the biggest risk,
|
|
208
|
+
stories that can't start, the oldest code review waiting past the pickup limit;
|
|
209
|
+
stories not ready to start take the place of an empty one, and an all-clear card
|
|
210
|
+
says so when nothing needs anyone; if a read failed, that card says which reads could not be read instead of an all-clear. A card's title is its one link, opening the item in
|
|
211
|
+
its drawer (or its page, as Customize says), with the ticket key in small print.
|
|
212
|
+
- **Outside the fold, too:** one compact line for each red reason no card already tells (it opens the reason in the fold), and the "Couldn't check" line for a panel that could not be read.
|
|
213
|
+
- **Show the full briefing**, closed until you open it, and remembered in this
|
|
214
|
+
browser: everything below, with Customize. The source and sprint line, any problem
|
|
215
|
+
with the briefing, and **Go to** stay in view outside it.
|
|
216
|
+
|
|
217
|
+
1. **The recommendation:** the one thing to do first, as a sentence. The sprint and
|
|
218
|
+
its day, and who wrote the briefing, stay under the answer, outside the fold: the latest Analyze, when it is
|
|
179
219
|
current (made since the data was read, or the same UTC day) and answered with a
|
|
180
220
|
briefing, with who wrote it, when, and a link to it; otherwise toil's briefing by
|
|
181
|
-
code (`toil
|
|
221
|
+
code (`toil projects brief`, `src/core/brief/`), with no agent call, which says so,
|
|
182
222
|
and why when an older Analyze exists.
|
|
183
223
|
2. **Why:** the reasons, one line each, the first open and the rest folded: Failing
|
|
184
224
|
checks, Blocked, Waiting on review, Sprint at risk, Stale, Over the WIP limit, and
|
|
185
|
-
Ticket and pull request disagree, in the
|
|
225
|
+
Ticket and pull request disagree, in the Toil project's order. Each item is under one
|
|
186
226
|
reason only, as a row with its key, title and assignee. Then **On track**, counted
|
|
187
227
|
rather than listed.
|
|
188
|
-
3. **One step in:** anything the
|
|
228
|
+
3. **One step in:** anything the Toil project's focus left out, with why, and anything the
|
|
189
229
|
code flagged that Analyze didn't mention, so nothing drops off the page unsaid. An
|
|
190
|
-
id Analyze named that isn't the
|
|
230
|
+
id Analyze named that isn't the Toil project's is shown in red.
|
|
191
231
|
4. **The numbers**, folded: burn-up, throughput, cycle time, aging work in progress
|
|
192
232
|
and pull request wait, each opening its full chart.
|
|
193
233
|
5. **Go to:** every other view, as well as its tab.
|
|
@@ -202,15 +242,15 @@ to the view behind it.
|
|
|
202
242
|
|
|
203
243
|
#### Tuning Analyze
|
|
204
244
|
|
|
205
|
-
Each
|
|
206
|
-
Analyze briefs** (`#/settings/
|
|
245
|
+
Each Toil project tunes how Analyze briefs it, under **Settings → Toil projects**, **How
|
|
246
|
+
Analyze briefs** (`#/settings/projects/<name>/analyze`): its goal, who it's written
|
|
207
247
|
for, its lenses in order with their thresholds, and what else to watch for or leave
|
|
208
248
|
out, the lenses folded until you open them. **The prompt Analyze will send**, under
|
|
209
249
|
the form, shows it as you type, composed by toil's own code, and **Save** runs
|
|
210
|
-
`toil
|
|
211
|
-
the form. The settings live in the
|
|
212
|
-
shares them; whenever a
|
|
213
|
-
whether it's your own, and a focus from a
|
|
250
|
+
`toil projects analyze set` with only what you changed, the line it runs shown under
|
|
251
|
+
the form. The settings live in the Toil project file, so a team that shares the Toil project
|
|
252
|
+
shares them; whenever a Toil project tunes Analyze, the briefing names the file and
|
|
253
|
+
whether it's your own, and a focus from a Toil project you haven't adopted still applies.
|
|
214
254
|
Clearing a setting is a delete, made in a terminal. Toil UI composes the preview with
|
|
215
255
|
the toil code it's built with, and Analyze asks the toil it runs on, so the two match
|
|
216
256
|
when Toil UI and toil are released together.
|
|
@@ -248,7 +288,7 @@ blocks another or is blocked, on the current sprint's dates, then the dependency
|
|
|
248
288
|
toil reads dates a link, so a block is drawn from when both items existed. An item
|
|
249
289
|
whose history records no move into In progress (GitHub records only open and closed)
|
|
250
290
|
is drawn in progress from when it was created. Nothing is fetched for the chart: it
|
|
251
|
-
reads the links, transitions, pull requests, builds and checks the
|
|
291
|
+
reads the links, transitions, pull requests, builds and checks the Toil project already
|
|
252
292
|
synced, and the readiness gate's verdict for work not started.
|
|
253
293
|
- **The table under the chart lists every bar and every arrow**, with links: each
|
|
254
294
|
item's status and points, its time in progress, blocked (and by what) and with
|
|
@@ -275,25 +315,27 @@ deployments themselves.
|
|
|
275
315
|
The board's cards and its **Needs attention** flag and the flow charts mark
|
|
276
316
|
each open work item with its insights, weightiest first, each naming the data it
|
|
277
317
|
rests on (`ui/src/model/insights.ts`). One the data can't support isn't raised at
|
|
278
|
-
all, rather than raised on a guess.
|
|
318
|
+
all, rather than raised on a guess. A card shows one chip per kind of insight,
|
|
319
|
+
with a count when there are several ("Blocking 2"); its tooltip lists each one,
|
|
320
|
+
and the item view lists them all.
|
|
279
321
|
|
|
280
322
|
| Insight | Means | Weight |
|
|
281
323
|
|---|---|---|
|
|
282
324
|
| Blocked | Another item that isn't done blocks it (its links) | needs action now |
|
|
283
325
|
| Build failing | Its pull request's latest build failed | needs action now |
|
|
284
|
-
| Aging | In progress longer than 85% of the
|
|
326
|
+
| Aging | In progress longer than 85% of the Toil project's finished items took; needs at least three cycle times to compare with | needs action now |
|
|
285
327
|
| Deployed, still open | An environment runs it, but it isn't done | needs a look |
|
|
286
328
|
| Carried over | It was in the previous sprint too | needs a look |
|
|
287
329
|
| PR with no build | A pull request for it is open with no build | needs a look |
|
|
288
330
|
| Stale | No update for more than 5 days | needs a look |
|
|
289
331
|
| Blocking | It blocks another item | needs a look |
|
|
290
|
-
| Merged, not deployed | Its pull requests are merged, but no environment runs them yet; only for a
|
|
332
|
+
| Merged, not deployed | Its pull requests are merged, but no environment runs them yet; only for a Toil project that names environments | worth knowing |
|
|
291
333
|
| No estimate | It has no estimate | worth knowing |
|
|
292
334
|
| Added mid-sprint | It joined the sprint after it started | worth knowing |
|
|
293
335
|
|
|
294
336
|
## Pull requests
|
|
295
337
|
|
|
296
|
-
Each
|
|
338
|
+
Each Toil project has a **Pull requests** tab (`#/project/<name>/prs`), built from OWI 0.2
|
|
297
339
|
changes:
|
|
298
340
|
|
|
299
341
|
- **Waiting now:** the open pull requests, longest wait first, each with what it
|
|
@@ -307,35 +349,36 @@ changes:
|
|
|
307
349
|
merge (approval to merge) times, review rounds, sizes and merges per week, over six
|
|
308
350
|
weeks, each split by people, agents and bots and each with its sample size and what
|
|
309
351
|
couldn't be measured.
|
|
352
|
+
- **Actions:** each open pull request has one **Actions** menu with **Review** (reads the pull request, its diff and checks, and drafts findings as comments you can post) and **Ship** (on a pull request a person wrote: drafts replies to its review; editing, the gate and pushing stay in your terminal). Each opens its own page, `#/project/<name>/job/review?target=<change id>` or `/job/ship?…`, which says what the job does and needs and has **Run**. The run goes through the Ask bar, so every write is a proposal card and the run is kept in the worklog. A job missing its code host's connection stays in the menu, dimmed, naming the connection to add. Ship stops after 3 finished rounds on one pull request; the next run is refused and says it needs a person (`needs-human`). The steps come from toil's workflow pages (`skills/toil/workflows/`).
|
|
310
353
|
- **Suggested reviewers:** for an open pull request with nobody asked, the CODEOWNERS
|
|
311
354
|
of its paths and the earlier reviewers of the same files and folders, each as a
|
|
312
355
|
proposal with the `toil … --allow-write` line to ask them.
|
|
313
356
|
|
|
314
|
-
The
|
|
357
|
+
The Toil project's work item keys are matched in each pull request's title, branch and
|
|
315
358
|
description (a Jira project's `PAY-101`, Azure Boards' `AB#2302`, a GitHub
|
|
316
359
|
repository's `owner/repo#41`). A code host's section may set author kinds by hand,
|
|
317
360
|
`"authorKinds": { "release-bot": "BOT" }`, which win over every other signal.
|
|
318
361
|
|
|
319
362
|
## People
|
|
320
363
|
|
|
321
|
-
Each
|
|
364
|
+
Each Toil project has a **People** tab (`#/project/<name>/people`), built
|
|
322
365
|
from the same OWI work items and changes as the dashboard and the pull requests view,
|
|
323
|
-
and from the
|
|
324
|
-
system (`toil
|
|
366
|
+
and from the Toil project's `people` section, which says who each person is in every
|
|
367
|
+
system (`toil projects person`, or **Settings, Toil projects**):
|
|
325
368
|
|
|
326
369
|
- **A card per person**, agents included, in name order: their work in progress by
|
|
327
370
|
status with how long each item has sat in it, their open pull requests and what
|
|
328
371
|
each waits on, the reviews waiting on them (oldest first), their blocked items and
|
|
329
372
|
what blocks each, and what they did in the last 14 days. A link opens their items
|
|
330
373
|
on the Board tab, filtered to them.
|
|
331
|
-
- **A team strip:** work in progress against the
|
|
374
|
+
- **A team strip:** work in progress against the Toil project's `wipLimit` and who is over
|
|
332
375
|
it, the review queue, and the current sprint's unassigned items.
|
|
333
376
|
- **Agents:** a person with `kind: "agent"`, matched by the same signals as a pull
|
|
334
377
|
request's author kind (its account, a Co-Authored-By trailer, a branch prefix) where
|
|
335
378
|
the signal names it. A person's own pull request written with an agent says so.
|
|
336
|
-
- **Accounts the
|
|
379
|
+
- **Accounts the Toil project doesn't map** get cards of their own, so no work goes
|
|
337
380
|
missing. One that shares an email with a person, or with an account in another
|
|
338
|
-
system, is suggested as a match with the `toil
|
|
381
|
+
system, is suggested as a match with the `toil projects person` line that applies
|
|
339
382
|
it; it is applied only when someone presses the button, and an email that fits more
|
|
340
383
|
than one candidate suggests nothing and says why. Accounts that only wrote bots'
|
|
341
384
|
pull requests are left out and counted.
|
|
@@ -348,20 +391,20 @@ stay team-level, in the pull requests view.
|
|
|
348
391
|
|
|
349
392
|
Before a work item goes to an agent, Toil UI checks whether it is ready for one. The
|
|
350
393
|
check uses no model and no premium request: only the item's cached OWI record, the
|
|
351
|
-
|
|
394
|
+
Toil project's other items, and at most one live read, made when you press for it
|
|
352
395
|
(`ui/src/model/readiness.ts`). Its verdict is **ready**, **thin** or **blocked**, with
|
|
353
396
|
one row per check:
|
|
354
397
|
|
|
355
398
|
| Check | Passes when | Otherwise |
|
|
356
399
|
|---|---|---|
|
|
357
|
-
| Acceptance criteria | The field the
|
|
358
|
-
| Description | At least `minWords` words (25 by default) of real text, not counting the
|
|
359
|
-
| Done state | It isn't done, and a way to Done is known: a GitHub issue closes from any state; otherwise an item of the same type in the
|
|
360
|
-
| Blockers | No open item blocks it | blocked by an open item; thin for a blocker the
|
|
361
|
-
| Scope | It comes from one of the
|
|
400
|
+
| Acceptance criteria | The field the Toil project names has them, or the description has an "Acceptance criteria", "Definition of done" or "Done when" heading with text under it, or a checklist | thin |
|
|
401
|
+
| Description | At least `minWords` words (25 by default) of real text, not counting the Toil project's template lines, placeholder lines (TBD, N/A, lorem ipsum) or unfilled `<slots>` | thin |
|
|
402
|
+
| Done state | It isn't done, and a way to Done is known: a GitHub issue closes from any state; otherwise an item of the same type in the Toil project reached Done, or the live read finds a Done state (Azure DevOps) or a move into Done (Jira) | blocked when it's already done, or the workflow has no Done state; thin when it isn't known |
|
|
403
|
+
| Blockers | No open item blocks it | blocked by an open item; thin for a blocker the Toil project doesn't hold |
|
|
404
|
+
| Scope | It comes from one of the Toil project's sources, and the Toil project names a repository | blocked from outside the Toil project; thin with no repository |
|
|
362
405
|
|
|
363
406
|
The verdict is blocked when any row is, thin when any row is, and ready only when every
|
|
364
|
-
row passes. Each area's section in the
|
|
407
|
+
row passes. Each area's section in the Toil project sets it up:
|
|
365
408
|
|
|
366
409
|
```json
|
|
367
410
|
"jira": {
|
|
@@ -378,10 +421,10 @@ that names none uses `Microsoft.VSTS.Common.AcceptanceCriteria`.
|
|
|
378
421
|
rows, each with where its answer came from. **Check the workflow live** makes the one
|
|
379
422
|
live read, read-only: `jira issue transitions` for a Jira issue, `ado work-item
|
|
380
423
|
states` for an Azure DevOps type. A GitHub issue needs none.
|
|
381
|
-
- **The Command Center** (`#/
|
|
424
|
+
- **The Command Center** (`#/project/<name>/command-center`, the Board tab's Readiness view) lists every
|
|
382
425
|
work item with its verdict and, for each one that isn't ready, the rows that say why.
|
|
383
426
|
It filters by verdict, status, type, assignee and source, and makes no live read.
|
|
384
|
-
It is strict: only ready items are slated, in the
|
|
427
|
+
It is strict: only ready items are slated, in the Toil project's order, under a plan id
|
|
385
428
|
that changes when anything in the plan does.
|
|
386
429
|
- **Starting the plan comes with assign-to-agent (#293).** Until then the start button
|
|
387
430
|
is disabled and says so; the plan it would start is shown above it.
|
|
@@ -389,10 +432,10 @@ that names none uses `Microsoft.VSTS.Common.AcceptanceCriteria`.
|
|
|
389
432
|
are ready for an agent?" (The agent, below).
|
|
390
433
|
## Worklog
|
|
391
434
|
|
|
392
|
-
Each
|
|
435
|
+
Each Toil project has a **Worklog** tab (`#/project/<name>/worklog`):
|
|
393
436
|
one page a day, folded by code from what the cache holds, facts only. No model writes
|
|
394
437
|
a line of it. A calendar shows the days that have a page, with how many lines each
|
|
395
|
-
has; a day opens as `#/
|
|
438
|
+
has; a day opens as `#/project/<name>/worklog/<yyyy-mm-dd>`. Days and times are UTC,
|
|
396
439
|
as everywhere else in Toil UI.
|
|
397
440
|
|
|
398
441
|
A page has:
|
|
@@ -435,8 +478,8 @@ page is folded again from the cache and written where its text changed, as Markd
|
|
|
435
478
|
in the data folder:
|
|
436
479
|
|
|
437
480
|
```text
|
|
438
|
-
<data folder>/cache/worklog/<
|
|
439
|
-
<data folder>/cache/worklog/<
|
|
481
|
+
<data folder>/cache/worklog/<toilProject>/<yyyy-mm>/<yyyy-mm-dd>.md
|
|
482
|
+
<data folder>/cache/worklog/<toilProject>/runs/<run id>.md
|
|
440
483
|
```
|
|
441
484
|
|
|
442
485
|
The pages live under the cache's retention: a day whose rows are gone loses its page,
|
|
@@ -447,16 +490,16 @@ it came out the same as the page kept; the same cache always gives the same page
|
|
|
447
490
|
toil's own file service: `toil util encode base64 --decode --out <file>`, run with that
|
|
448
491
|
folder as its workspace, which writes a new file only and never into toil's own
|
|
449
492
|
config. Publishing an exported page is a proposal: `toil confluence page create
|
|
450
|
-
--body-file <file> --markdown` in the
|
|
493
|
+
--body-file <file> --markdown` in the Toil project's Confluence space, or `toil github issue
|
|
451
494
|
create --body-file <file>` in its repository, for you to run through toil's write gate.
|
|
452
495
|
Toil UI posts nothing. With sample data, notes and export are off.
|
|
453
496
|
|
|
454
497
|
### Stand-up
|
|
455
498
|
|
|
456
499
|
**Stand-up since the last working day**, on the Worklog
|
|
457
|
-
(`#/
|
|
500
|
+
(`#/project/<name>/worklog/standup`), reshapes the worklog for a meeting: since the last
|
|
458
501
|
working day through today (Friday through Monday on a Monday), or any range of up to 31
|
|
459
|
-
days you pick (`#/
|
|
502
|
+
days you pick (`#/project/<name>/worklog/standup/<from>/<to>`, or `/<day>` for one), and each day's page has a
|
|
460
503
|
**Stand-up for this day**. It is a fold of those days' pages by code, with no model, so the
|
|
461
504
|
same days always give the same report.
|
|
462
505
|
|
|
@@ -468,9 +511,9 @@ same days always give the same report.
|
|
|
468
511
|
failed build of their pull request, a pull request of theirs waiting on review for 2
|
|
469
512
|
working days or more) and **Next** (current-sprint items assigned to them and not
|
|
470
513
|
started). A line belongs to whoever the fact names: who moved the item, who wrote or
|
|
471
|
-
reviewed the pull request. A pull request a
|
|
472
|
-
- **People** come from the
|
|
473
|
-
GitHub accounts land under one name. An account the
|
|
514
|
+
reviewed the pull request. A pull request a Toil project agent wrote goes under the agent.
|
|
515
|
+
- **People** come from the Toil project's `people`, so one person's Jira, Azure DevOps and
|
|
516
|
+
GitHub accounts land under one name. An account the Toil project doesn't map gets its own
|
|
474
517
|
card, named with its area and id; the People view suggests where it belongs.
|
|
475
518
|
- **What the worklog can't vouch for** in the range is said at the top. Review waits, the
|
|
476
519
|
sprint and Next aren't on any worklog page, so they are as of the latest sync, and the
|
|
@@ -479,7 +522,7 @@ same days always give the same report.
|
|
|
479
522
|
|
|
480
523
|
It copies as **Markdown** (for Slack, Teams or a wiki) or **plain text**, line for line as
|
|
481
524
|
shown. **Export** writes the Markdown through toil's file service, as a page's export does,
|
|
482
|
-
to `standup-<
|
|
525
|
+
to `standup-<toilProject>-<from>-to-<to>.md`, and publishing it to Confluence or GitHub is a
|
|
483
526
|
proposal for you to run.
|
|
484
527
|
|
|
485
528
|
**Write it up** is optional: it asks the agent for a short narrative of the same report,
|
|
@@ -490,7 +533,7 @@ usage counts like any chat's.
|
|
|
490
533
|
|
|
491
534
|
## Toil's configuration
|
|
492
535
|
|
|
493
|
-
The Settings pages manage what `toil connections` and `toil
|
|
536
|
+
The Settings pages manage what `toil connections` and `toil projects` manage, by
|
|
494
537
|
running those commands through `@imagile-dev/toil/api`, never by editing a file:
|
|
495
538
|
|
|
496
539
|
- **Connections:** list, add, test, set the default, turn writes on or off,
|
|
@@ -505,9 +548,9 @@ running those commands through `@imagile-dev/toil/api`, never by editing a file:
|
|
|
505
548
|
Turning writes on asks you to type the connection's name, as the CLI does. A
|
|
506
549
|
connection made by `TOIL_*` env vars alone is shown read-only, with where it
|
|
507
550
|
came from.
|
|
508
|
-
- **
|
|
509
|
-
show and check any of them. A
|
|
510
|
-
on its card and dashboard, through `toil
|
|
551
|
+
- **Toil projects:** create one, add a repository, link an area, name a field, and
|
|
552
|
+
show and check any of them. A Toil project's name and description are edited in place
|
|
553
|
+
on its card and dashboard, through `toil projects describe`. A link to a connection with writes on is refused
|
|
511
554
|
here, as over MCP: make it on the CLI.
|
|
512
555
|
- **Bounds and time limit:** toil has no command for these, so they are shown
|
|
513
556
|
read-only, with the file or env var that sets them.
|
|
@@ -529,7 +572,7 @@ newer than 0.1.0-alpha.2; with that one, use the CLI.
|
|
|
529
572
|
An **(i)** beside a term explains it in a sentence or two when you hover or focus
|
|
530
573
|
it, and pressing it opens the same words with a **Read more** link to the section
|
|
531
574
|
of these docs that says the rest. A **pencil** beside a configurable value (a
|
|
532
|
-
connection, a
|
|
575
|
+
connection, a Toil project's sources, people, field names or work-in-progress limit,
|
|
533
576
|
the agent, a usage cap) opens the settings page where it's set, at that value; it
|
|
534
577
|
only navigates. **Help** in the header is a menu with the docs for the view you're on
|
|
535
578
|
and **What's new** (below).
|
|
@@ -540,7 +583,7 @@ dialog's modes, session and system prompt in Renew all stale, hallucination in t
|
|
|
540
583
|
chat's **About these answers**), its (i) popover also gives each term's first
|
|
541
584
|
sentence from Context Overflow's glossary, with **Read more on Context Overflow**.
|
|
542
585
|
Help's menu then has **Learn**: four of Context Overflow's modules, by task.
|
|
543
|
-
|
|
586
|
+
**Links to learning content**, under
|
|
544
587
|
**Settings, General** and kept in this browser, is on by default; off, every link to
|
|
545
588
|
Context Overflow and every glossary line is gone, and Help keeps the view's docs and
|
|
546
589
|
What's new.
|
|
@@ -554,12 +597,12 @@ views follow.
|
|
|
554
597
|
|
|
555
598
|
Every empty panel offers the one action that fills it, under one short line, with any
|
|
556
599
|
explanation in its (i): **Link a pipeline** and **Add an environment** (pencils to
|
|
557
|
-
|
|
600
|
+
Toil project setup at the Toil project's sources) on the Board tab's builds and deployments, **Add
|
|
558
601
|
a repository** for pull requests, **Add a person** on the People tab, **Analyze now**
|
|
559
602
|
for an empty Analyses tab, **Show every item** for a Command Center filter that hides
|
|
560
603
|
everything, and **Sync now** in the local app wherever only a newer read of the tracker
|
|
561
604
|
fills the panel (the sprint, the burn-up, links between items). The site preview can't
|
|
562
|
-
sync, so it offers
|
|
605
|
+
sync, so it offers Toil project setup there instead. A test fails on any empty panel with
|
|
563
606
|
neither an action nor a link.
|
|
564
607
|
|
|
565
608
|
### What's new
|
|
@@ -580,23 +623,23 @@ screen it's **Search and commands** in the header's Menu. Type to narrow it down
|
|
|
580
623
|
use the arrow keys to move, **Enter** to open and **Escape** to close. It's hidden
|
|
581
624
|
until you call it up.
|
|
582
625
|
|
|
583
|
-
- **Jump** to a
|
|
626
|
+
- **Jump** to a Toil project, any of a Toil project's tabs (Overview, Board, Dependencies,
|
|
584
627
|
Pull requests, People, Worklog, Analyses), a saved analysis, a settings page, or a
|
|
585
628
|
theme.
|
|
586
|
-
- **Run** Sync now (where Toil UI keeps a cache), and on a
|
|
587
|
-
focuses the Ask bar), Analyze, the agent that answers this chat, and Assign the
|
|
629
|
+
- **Run** Sync now (where Toil UI keeps a cache), **Use plain words** / **Use toil's own terms**, and on a Toil project's page Ask (it
|
|
630
|
+
focuses the Ask bar), Analyze, **Review a pull request**, **Ship: answer a review** and **Diagram <Toil project>** (each opens its job's page), **Open the Agents panel** (**Open the AI helpers panel** in plain words), the agent that answers this chat, and Assign the
|
|
588
631
|
item you were on to an agent: the board card that had focus, or the item whose
|
|
589
632
|
page is open.
|
|
590
633
|
- Matching is fuzzy and ignores case: `pay12` finds `PAY-12`, and each word you
|
|
591
634
|
type has to appear, in order. Results are grouped by kind, and the group with the
|
|
592
635
|
best match comes first.
|
|
593
636
|
|
|
594
|
-
### Search across
|
|
637
|
+
### Search across Toil projects
|
|
595
638
|
|
|
596
|
-
With something typed, the palette searches the items of every
|
|
597
|
-
title. Each result shows its
|
|
598
|
-
what the last sync stored, never the tracker: opening it reads each
|
|
599
|
-
the cache the way its dashboard does. A result from a
|
|
639
|
+
With something typed, the palette searches the items of every Toil project, by key and
|
|
640
|
+
title. Each result shows its Toil project and status and opens the item. It reads only
|
|
641
|
+
what the last sync stored, never the tracker: opening it reads each Toil project from
|
|
642
|
+
the cache the way its dashboard does. A result from a Toil project last synced more than
|
|
600
643
|
a day ago, or whose last sync failed, says how old its copy is.
|
|
601
644
|
|
|
602
645
|
### Keyboard shortcuts
|
|
@@ -621,7 +664,7 @@ request: the kind, a title, the details and your email. It sends through
|
|
|
621
664
|
first. Send works only while the box is ticked, and any edit clears it, so you
|
|
622
665
|
agree to the report on screen and no other.
|
|
623
666
|
- **Only what you typed.** toil adds its version and commit, nothing else: no
|
|
624
|
-
|
|
667
|
+
Toil project, connection, host or cached data.
|
|
625
668
|
- **The same lock as Settings.** `api/feedback` needs the session, the page's
|
|
626
669
|
Origin and the CSRF token, takes the four fields and nothing else, and runs
|
|
627
670
|
`toil feedback` and no other command. Your tick becomes the yes to the command's
|
|
@@ -633,7 +676,7 @@ request: the kind, a title, the details and your email. It sends through
|
|
|
633
676
|
|
|
634
677
|
The dashboards read from a local cache, which a sync fills. Press **Sync now** in
|
|
635
678
|
the Sync panel at the foot of the page. The sync makes the same toil calls the
|
|
636
|
-
dashboards make, for every
|
|
679
|
+
dashboards make, for every Toil project, and stores the answers. The work items go
|
|
637
680
|
through toil's Open Work Item mapping, with their status changes and the pull
|
|
638
681
|
requests, builds and deployments that name them.
|
|
639
682
|
|
|
@@ -646,27 +689,27 @@ requests, builds and deployments that name them.
|
|
|
646
689
|
Once a day, the Jira and Azure DevOps searches are also sent whole (ids only
|
|
647
690
|
where toil can), and an item they no longer list is dropped; if that listing
|
|
648
691
|
fails, the changes are kept and the next sync tries it again. Per-item calls
|
|
649
|
-
(histories, pull request gets) run 4 at a time across a
|
|
692
|
+
(histories, pull request gets) run 4 at a time across a Toil project.
|
|
650
693
|
Azure DevOps histories come from the project's revisions feed, read from the
|
|
651
694
|
watermark; an item is read on its own when the feed doesn't hold all of it.
|
|
652
695
|
- **Every page.** Each call follows `meta.page.nextCursor` to the end, up to 200
|
|
653
696
|
pages, and fails that source, saying so, rather than keep part of a listing.
|
|
654
|
-
- **Each source on its own.** Each area a
|
|
697
|
+
- **Each source on its own.** Each area a Toil project reads (Jira, Bitbucket, Jenkins,
|
|
655
698
|
OpenShift and so on) is a source with its own state. A source that fails shows
|
|
656
699
|
toil's message and fix in the Sync panel, and keeps its last good data, marked
|
|
657
|
-
stale. The other sources and
|
|
700
|
+
stale. The other sources and Toil projects aren't touched. The next sync asks it
|
|
658
701
|
for everything since its last good sync, so nothing is missed.
|
|
659
702
|
- **When.** The line under the header says when the data was synced. The panel
|
|
660
|
-
says when each
|
|
703
|
+
says when each Toil project was synced, and shows a running sync's progress
|
|
661
704
|
(requests, pages and retries, from toil's own progress events).
|
|
662
705
|
- **Where.** The cache is one SQLite file, `cache/toil-ui.db` in the data folder,
|
|
663
706
|
through Node's built-in `node:sqlite`, so there's no native module. While Toil
|
|
664
707
|
UI runs, its write-ahead log (`toil-ui.db-wal` and `toil-ui.db-shm`) sits beside
|
|
665
708
|
it, and closing folds the log back into the file. It holds
|
|
666
709
|
ticket content, so it stays on this machine and is readable by your account
|
|
667
|
-
only. The analyses (#395) are kept in the same file, the newest 100 per
|
|
710
|
+
only. The analyses (#395) are kept in the same file, the newest 100 per Toil project,
|
|
668
711
|
and a sync never rewrites them.
|
|
669
|
-
- **For how long.** A
|
|
712
|
+
- **For how long.** A Toil project's synced data (responses, items,
|
|
670
713
|
transitions, change refs, worklog facts and each source's state) expires together, once the
|
|
671
714
|
last sync that saved data for it is older than the retention period (30 days by default),
|
|
672
715
|
at start and after each sync. A sync that fails doesn't keep old data alive. Worklog facts that no
|
|
@@ -674,7 +717,7 @@ requests, builds and deployments that name them.
|
|
|
674
717
|
goes when it is that old itself. The retention period
|
|
675
718
|
also sets how far back closed Jira and Azure DevOps items are read, with a
|
|
676
719
|
floor of 6 weeks (see What it reads). An analysis goes once
|
|
677
|
-
it is older than the retention period, and with its
|
|
720
|
+
it is older than the retention period, and with its Toil project when toil no longer
|
|
678
721
|
has it.
|
|
679
722
|
- **Wiping it.** **Wipe the cache** in the Sync panel deletes it after asking
|
|
680
723
|
first. `toil-ui wipe` does the same from a terminal, once `toil-ui` has
|
|
@@ -686,7 +729,7 @@ requests, builds and deployments that name them.
|
|
|
686
729
|
exist. So is an item closed longer ago than the read window. A GitHub issue
|
|
687
730
|
stays until the retention period ends or you wipe the cache.
|
|
688
731
|
|
|
689
|
-
The page can run only the reads the dashboards make: `
|
|
732
|
+
The page can run only the reads the dashboards make: `projects list` and `Toil projects
|
|
690
733
|
show`, and each area's search, history, pull request, build and deployment reads.
|
|
691
734
|
Any other command is refused before it reaches toil, and a change to a tracker is
|
|
692
735
|
never made from the page. The worklog's own calls are made by the server, never named
|
|
@@ -695,39 +738,61 @@ by the page: `toil audit list`, read-only, to see whether a proposal was applied
|
|
|
695
738
|
|
|
696
739
|
## The agent
|
|
697
740
|
|
|
698
|
-
Every
|
|
699
|
-
|
|
700
|
-
screen and the page steps back, with a slim strip that keeps the dossier, its sprint
|
|
741
|
+
Every Toil project page has an **Ask bar** in its top row: one line. Pressing into it opens the **focused view**: the conversation takes the
|
|
742
|
+
screen and the page steps back, with a slim strip that keeps the Toil project, its sprint
|
|
701
743
|
and the agent answering in sight. **Escape** or **Close** puts it away; the
|
|
702
|
-
conversation stays, for the rest of the
|
|
703
|
-
|
|
704
|
-
- **
|
|
705
|
-
|
|
744
|
+
conversation stays, for the rest of the Toil project's pages, until you reload.
|
|
745
|
+
|
|
746
|
+
- **The guided start** (#552) is what the chat shows before the first question:
|
|
747
|
+
"What would you like to know about {name}?", then nine plain questions in four groups
|
|
748
|
+
(How are we doing? What's in the way? Backlog. For others), beside **Before you ask**. What each is good for is its hover; the one line on the page is that "Who might need help?" never ranks anyone. The dialog's title is the project's name. You can still type your own, except with no AI helper set up, when the questions are off and the notice links to **Settings → AI helper**. Pressing a question asks its text
|
|
749
|
+
through the same path as typing it; after the first answer the same questions are the
|
|
750
|
+
**Suggestions** menu. The questions are `src/guided-questions.ts`, not the server's
|
|
751
|
+
starters, which still back `api/agent/starters`.
|
|
752
|
+
- **Before you ask** is on the page, never in a popover, as the same one line at every width,
|
|
753
|
+
"It reads what your connections can read, saved and live. It changes nothing.", with a
|
|
754
|
+
**What AI can see** button that opens the rest: what the AI can see (this Toil project's saved
|
|
755
|
+
copy, counted as items, code reviews, builds and deployments, and live reads through toil when
|
|
756
|
+
the copy is not enough, read-only) and what it can't do (change anything in the systems named,
|
|
757
|
+
see email or chat, read anything your connections can't already read). Who answers is the
|
|
758
|
+
header's to say, and what it costs is the usage line's.
|
|
759
|
+
- **How sure is this?** is under every finished answer, worked out by code
|
|
760
|
+
(`src/model/confidence.ts`) and never by the model, with an (i). **High**: every saved read worked, the data
|
|
761
|
+
is under 24 hours old, and at most 1 in 10 of the items the question needs a field of
|
|
762
|
+
lacks it. **Medium**: one of those slipped, or the answer names one item its tools never returned. **Low**: a read failed, the data is over 3 days
|
|
763
|
+
old, there are fewer than 3 items to go on (code reviews, for the question about reviews), or the answer names more than one item its tools never returned, or half of those it cites. A question from the guided start is checked against
|
|
764
|
+
the panels it needs; a typed question and Analyze against every read. Under the word and its reason is one folded line, "What it looked at · 3 s · about $0.04 of usage (estimate)", which opens to the items the answer cited as links (a cited id the project doesn't hold as the bare
|
|
765
|
+
id) and the usage breakdown. "Suggestions only. Nothing was changed." is said under the first answer of a conversation, not every one; "Answered by" appears under an answer only when it differs from the header's. The answer's own heading is not drawn (it repeats the question), though a screen reader still has it. The existing proposal
|
|
766
|
+
cards still show the command to run, with no approve button.
|
|
706
767
|
- **Answers stream in** as the agent writes them, under a **status line** that says
|
|
707
768
|
what it is doing: reading the cache with `owi_query`, reading live through toil with
|
|
708
769
|
`github issue list`, writing, and at the end how long it took. **Stop** cancels the
|
|
709
770
|
question on the server and keeps what had arrived. Claude and Copilot stream as they
|
|
710
771
|
write; Codex's SDK hands over its text a message at a time, so with Codex the status
|
|
711
772
|
line is what moves between messages.
|
|
712
|
-
- **The wording follows the theme
|
|
713
|
-
words are in each game theme's voice, and Plain stays neutral. The starter questions
|
|
773
|
+
- **The wording follows the theme** when themed wording is on and plain words are off: the bar's placeholder, the welcome and the status
|
|
774
|
+
words are in each game theme's voice, and Plain stays neutral. With plain words on (the default), they read in plain words. The starter questions
|
|
714
775
|
read the same in every theme.
|
|
715
|
-
- **Analyze** briefs the
|
|
716
|
-
|
|
717
|
-
|
|
776
|
+
- **Analyze** briefs the Toil project in one press (#494). It is the guided start's first
|
|
777
|
+
question, "Give me the full briefing", and still the command palette's Get a briefing and the
|
|
778
|
+
Analyses tab's button; the Ask bar has no button of its own for it. Its question comes from toil,
|
|
779
|
+
`toil projects analyze prompt <toilProject> --host toil-ui`: the standard method with
|
|
780
|
+
the Toil project's own lenses and focus (Tuning Analyze, above). It answers with a
|
|
718
781
|
briefing block, which the server lifts out of the answer and checks, every id in it
|
|
719
782
|
against what the agent's tools read: a recommendation, then the reasons, each item
|
|
720
783
|
with why it's there and the next step, then what changed, and anything the
|
|
721
|
-
|
|
784
|
+
Toil project's focus left out. The briefing, the item drawer, the chat and the analysis
|
|
722
785
|
page all lay it out, with the rest of the answer as **Evidence**, folded until you
|
|
723
786
|
open it. An answer with no briefing block is shown as it came, and says so. With a
|
|
724
|
-
toil that has no `
|
|
787
|
+
toil that has no `projects analyze prompt`, Analyze asks Toil UI's own question
|
|
725
788
|
instead, and the answer warns that it did. Each whole Analyze is kept in the
|
|
726
|
-
|
|
727
|
-
- **Citations are
|
|
728
|
-
|
|
729
|
-
|
|
730
|
-
|
|
789
|
+
Toil project's **Analyses** tab, and the answer links to it.
|
|
790
|
+
- **Citations are links in the text**: in every answer, each `[[id]]` is an ordinary link
|
|
791
|
+
(a short source name such as ADO or GH, the item's key and a cut title) to its page (a
|
|
792
|
+
pull request to its host). Hovering or focusing one shows the whole title, its source and
|
|
793
|
+
status, type, assignee, estimate, sprint and last update; the status is also in the
|
|
794
|
+
link's name for a screen reader. An id the
|
|
795
|
+
Toil project doesn't hold stays the bare id, marked, and the server's check above the
|
|
731
796
|
answer still names any id its tools never returned.
|
|
732
797
|
- **Each question carries the chat so far**: its last 8 questions and answers, each
|
|
733
798
|
answer cut to 1,500 characters. An item an earlier answer named is still checked
|
|
@@ -737,16 +802,17 @@ conversation stays, for the rest of the dossier's pages, until you reload.
|
|
|
737
802
|
from its next question, which carries the conversation so far; the default stays in
|
|
738
803
|
Settings. An adapter that can't answer is listed with the reason and **Install** or
|
|
739
804
|
**Fix**, never left out. Each answer names the adapter and model that wrote it.
|
|
740
|
-
- **
|
|
741
|
-
agent it was asked of, in `<data folder>/
|
|
742
|
-
|
|
743
|
-
opens on that
|
|
744
|
-
default from Settings, Agent (Claude, GitHub Copilot or Codex), else
|
|
745
|
-
the server trying the default and saying why it
|
|
746
|
-
by** say which: "last used on this
|
|
747
|
-
|
|
748
|
-
|
|
749
|
-
|
|
805
|
+
- **The chat remembers its agent for each Toil project** (#494). A whole answer, Analyze or an
|
|
806
|
+
ordinary question, records the agent it was asked of, in `<data folder>/toil-project-agents.json` (an adapter name per
|
|
807
|
+
Toil project name, Toil UI's own state, never in the Toil project file). The next time the chat
|
|
808
|
+
opens on that Toil project it starts on that agent if it can still answer, else on the
|
|
809
|
+
default from Settings, Agent (Claude, GitHub Copilot or Codex), else the first agent
|
|
810
|
+
that is ready, else as before, with the server trying the default and saying why it
|
|
811
|
+
can't. A few words beside **Answered by** say which: "last used on this Toil project",
|
|
812
|
+
"your default", "your default can't run", "the first one ready" or
|
|
813
|
+
"picked for this chat". Picking another agent and asking updates the memory; a question that names no
|
|
814
|
+
agent, or an answer that fails, is stopped or comes back empty, doesn't. Renaming a Toil project moves its
|
|
815
|
+
memory, and a Toil project toil no longer has is forgotten on the next full sync. The sample
|
|
750
816
|
remembers nothing.
|
|
751
817
|
|
|
752
818
|
The chat answers with the agent your organization already approved, on the
|
|
@@ -766,9 +832,12 @@ toil's config, and has no field for a key: each SDK signs in the way its CLI doe
|
|
|
766
832
|
|
|
767
833
|
### Installing an adapter's SDK
|
|
768
834
|
|
|
769
|
-
**Settings, Agent**
|
|
770
|
-
|
|
771
|
-
|
|
835
|
+
**Settings, Agent** (**AI helper** in plain words) opens on the steps to get going and each
|
|
836
|
+
helper with its state (Ready, Not set up yet in a calm tone, or Needs one more step with the
|
|
837
|
+
reason shown), and an **Install** button on each helper that isn't installed. Whether its SDK
|
|
838
|
+
is installed and at which version against the one Toil UI was tested with, which CLI it would
|
|
839
|
+
start, and anything worth knowing, are one step in under **Details for whoever sets it up**;
|
|
840
|
+
models, the CLI path and the endpoint are behind **Advanced**. A missing SDK has an **Install** button, which shows the one
|
|
772
841
|
command it will run, word for word, before anything runs:
|
|
773
842
|
|
|
774
843
|
```bash
|
|
@@ -816,12 +885,12 @@ for (docs/decisions/343-ask-up-front.md).
|
|
|
816
885
|
`TOIL_*` variables, `GITHUB_TOKEN` and CI tokens never reach it. Not found is a plain
|
|
817
886
|
error naming the CLI and the fix. On Windows a `.cmd` shim can't be started by the
|
|
818
887
|
SDKs, so point at the `.exe` (Copilot's npm shim is the exception, above).
|
|
819
|
-
- **Read-only tools, one
|
|
888
|
+
- **Read-only tools, one Toil project.** The agent gets `owi_query`, `owi_get`,
|
|
820
889
|
`owi_similar` (likely duplicates, scored as `toil util owi similar` scores them)
|
|
821
890
|
and `owi_history` over the cache, `pr_query`, `pr_get` and `pr_metrics` over its
|
|
822
891
|
pull requests (what each waits on, its flags, and flow figures by author kind with
|
|
823
892
|
their sample sizes), `people_list` and `person_get` (each person's plate as the
|
|
824
|
-
People view works it out, with the
|
|
893
|
+
People view works it out, with the Toil project read live through `projects show`, and
|
|
825
894
|
told it describes load, never performance), `readiness_list` and
|
|
826
895
|
`readiness_get` (each item's readiness verdict and its rows, as the Command Center
|
|
827
896
|
works them out), `proposal_fields` (a write command's input fields,
|
|
@@ -829,7 +898,7 @@ for (docs/decisions/343-ask-up-front.md).
|
|
|
829
898
|
with what it can't vouch for), `standup_report` (a range of days as the Stand-up folds
|
|
830
899
|
them, each line numbered), `proposal_fields` (a write command's input fields,
|
|
831
900
|
from toil's registry), and `toil mcp`'s read tools for live lookups, called through
|
|
832
|
-
toil's `execute()` with `readOnly: true`. The
|
|
901
|
+
toil's `execute()` with `readOnly: true`. The Toil project is fixed to the one on screen.
|
|
833
902
|
There is no write tool, and each SDK's own tools (shell, files, web, plugins, other
|
|
834
903
|
MCP servers) are switched off.
|
|
835
904
|
- **Citations are checked.** Every answer cites the work items and pull requests it
|
|
@@ -863,18 +932,23 @@ for (docs/decisions/343-ask-up-front.md).
|
|
|
863
932
|
proposals or what the tools can do. **Show the full prompt** shows exactly what
|
|
864
933
|
the agent is sent.
|
|
865
934
|
|
|
866
|
-
With no adapter chosen, the chat says
|
|
867
|
-
|
|
868
|
-
|
|
935
|
+
With no adapter chosen, the chat says in plain words that no AI helper is set up and where
|
|
936
|
+
an engineer sets one up (Settings, Agent), with the three adapters and their packages one
|
|
937
|
+
step in, behind **Details for whoever sets it up**; its switcher can still
|
|
938
|
+
choose an installed one for that chat. The notice goes as soon as a helper can answer: the chat
|
|
939
|
+
reads the helpers again whenever the page moves, so installing on the Settings page needs no
|
|
940
|
+
reload. A helper that is installed but can't start yet says so instead, with the reason one step
|
|
941
|
+
in. `toil-ui --sample` keeps the scripted agent,
|
|
942
|
+
which answers its starter questions, Analyze and the workflow jobs from the sample data (a job's answer says nothing was read and lists every check as unchecked).
|
|
869
943
|
|
|
870
944
|
### Not ticketed yet
|
|
871
945
|
|
|
872
946
|
When you ask the chat for work that names no ticket (not a question, which ends in `?` or
|
|
873
947
|
opens with what, why, how and the like, and no key such as `PAY-12` or `#42`), a **Not ticketed yet** banner lists up to three of the
|
|
874
|
-
|
|
948
|
+
Toil project's open items likeliest to cover it, scored as `toil util owi similar` scores
|
|
875
949
|
them, each with why. Attach the work to one by naming its key in your branch and
|
|
876
950
|
commits, which writes nothing, or press **Ticket this** for proposals: a sub-task
|
|
877
|
-
under each match, and a new ticket on the
|
|
951
|
+
under each match, and a new ticket on the Toil project's tracker. They are proposal cards
|
|
878
952
|
like the chat's others, with the command line to run through toil's write gate, and
|
|
879
953
|
nothing is created until you run one. With no match, it says there is no likely
|
|
880
954
|
ticket and offers only a new one. It asks once per chat, and **Not now** hides it.
|
|
@@ -891,20 +965,20 @@ called):
|
|
|
891
965
|
| Codex | `turn.completed`'s `usage`: input (which holds the cached input), cached input, cache writes, output | tokens; a dollar figure only from rates you enter, labelled "computed from your rates". The SDK names no model, so a Codex record names the model you chose, or says it wasn't named |
|
|
892
966
|
| GitHub Copilot | one `assistant.usage` event per model call, with `cost` as the premium-request multiplier, and `session.usage_info` for the context. Copilot bills the multiplier once per prompt: only the call the prompt started (`initiator` `user`) counts it, and the agent's follow-up calls in the tool loop count 0 | tokens and premium requests, never dollars; premium requests left from the SDK's experimental `account.getQuota`, marked experimental |
|
|
893
967
|
|
|
894
|
-
- **Under each answer**, one line per model: `2.1k in · 380 out · 14k cached · cache
|
|
968
|
+
- **Under each answer**, one line in its folded line, "about $0.04 of usage (estimate)", or "(not billed on your plan)" on a Claude plan. The fold opens to one line per model: `2.1k in · 380 out · 14k cached · cache
|
|
895
969
|
write not reported · ~$0.012 (SDK estimate)`. A figure the SDK didn't report reads
|
|
896
|
-
"not reported", never 0.
|
|
897
|
-
- **In the chat's
|
|
970
|
+
"not reported", never 0. A cap notice stays in the open.
|
|
971
|
+
- **In the chat's footer**, from the second answer (one answer's usage is the chat's), the chat's running total, split by model, and how full
|
|
898
972
|
the last answer's context got, where the SDK says (Claude and Copilot; Codex's SDK
|
|
899
|
-
doesn't).
|
|
973
|
+
doesn't). On a phone the chat's header is one line, "Answered by … · Change".
|
|
900
974
|
- **Settings, Agent usage** (beside Usage, which counts toil's own commands): the last
|
|
901
|
-
30 days by day, adapter, model and
|
|
975
|
+
30 days by day, adapter, model and Toil project, the chats that used the most, and an
|
|
902
976
|
**Export as CSV** download made in the page.
|
|
903
977
|
- **One adding-up rule**, `addUp` in `src/model-usage.ts`, read by the server's caps,
|
|
904
978
|
the footer, the heading and Settings: reports with one id are one report, merged to
|
|
905
979
|
the highest of each figure (a running total), and different ids are summed.
|
|
906
980
|
- **The ledger** is the cache's `usage` table, apart from everything a sync writes: a
|
|
907
|
-
sync's cleanup or a
|
|
981
|
+
sync's cleanup or a Toil project leaving never lowers a total. It follows the cache's
|
|
908
982
|
retention, `toil-ui wipe` clears it, and it is never sent anywhere.
|
|
909
983
|
|
|
910
984
|
**Caps** are optional and off until you turn one on, under Settings, Agent usage: a
|
|
@@ -930,29 +1004,29 @@ you apply it.
|
|
|
930
1004
|
|---|---|---|
|
|
931
1005
|
| **Refine** | Checks for duplicates, drafts acceptance criteria, suggests an estimate, splits the item into children where it is too big, and asks its open questions | Nothing: every change is a proposal |
|
|
932
1006
|
| **Investigate** | Reads the item, its parent, children, links, likely duplicates, pull requests and toil's live reads, and writes a report that cites each source | Nothing: the report is kept with the assignment, and posting it as a comment is a proposal |
|
|
933
|
-
| **Implement** | Works in a copy of the
|
|
1007
|
+
| **Implement** | Works in a copy of the Toil project repository's git folder that Toil UI makes in a folder of its own, fills it from the last commit, makes the branch `agent/<key>`, changes and tests the code, and stages the change | Only that clone, with no network. Committing, pushing and opening a pull request are proposals |
|
|
934
1008
|
|
|
935
1009
|
- **Refine and investigate** get the same read-only tools as the chat (`owi_*`, `pr_*`,
|
|
936
1010
|
`proposal_fields` and toil's live reads), with no shell and no file writes.
|
|
937
1011
|
- **Implement runs on Codex only, for now.** Codex's own sandbox, `workspace-write`,
|
|
938
1012
|
holds the agent's commands and writes to the clone, with no network and none of the
|
|
939
1013
|
temp folders Codex adds by default, on Windows (its `unelevated` restricted-token sandbox), macOS and Linux. The clone is
|
|
940
|
-
`<data folder>/cache/work/<assignment>/repo`, so `toil-ui wipe` deletes it; before any git runs there outside the sandbox, Toil UI rewrites its git config and removes its hooks. The
|
|
941
|
-
repository needs a local path (`toil
|
|
1014
|
+
`<data folder>/cache/work/<assignment>/repo`, so `toil-ui wipe` deletes it; before any git runs there outside the sandbox, Toil UI rewrites its git config and removes its hooks. The Toil project's
|
|
1015
|
+
repository needs a local path (`toil projects add-repo <toilProject> <folder>`). The agent stages its change and never commits, pushes or fetches; the
|
|
942
1016
|
staged diff is shown through toil's own `git status` and `git diff`, and the commit
|
|
943
1017
|
and push are a proposal of the exact git lines to run, since toil has no command
|
|
944
1018
|
that pushes.
|
|
945
1019
|
- **States**: `queued` (waiting for a free slot, or for a person to press Start),
|
|
946
1020
|
`working`, `needs you` (proposals or questions to review, or an item the agent
|
|
947
1021
|
thought was trying to instruct it), `done` and `failed`. The **Agents** sidebar beside
|
|
948
|
-
every
|
|
1022
|
+
every Toil project tab shows each assignment's state, mode, adapter, run time and what it
|
|
949
1023
|
used, and the People view shows each agent's assignments.
|
|
950
1024
|
- **From the Command Center**, "Start every ready item" assigns each slated item in
|
|
951
1025
|
implement mode, in order, on Codex (or the adapter chosen in Settings, when it can
|
|
952
|
-
implement), in the
|
|
1026
|
+
implement), in the Toil project's one repository with a local path, and says what it
|
|
953
1027
|
queued or why it started nothing.
|
|
954
1028
|
- **The tracker only queues.** An item labelled `agent:refine`, `agent:investigate` or
|
|
955
|
-
`agent:implement`, or assigned to one of the
|
|
1029
|
+
`agent:implement`, or assigned to one of the Toil project's agents (its `people` with
|
|
956
1030
|
`kind: agent`, which queues a refine), appears in the **Agent inbox** on the next
|
|
957
1031
|
read. It runs only when a person presses Start.
|
|
958
1032
|
- **Applying a proposal** runs its toil command, the one the server stored when the
|
|
@@ -969,7 +1043,7 @@ you apply it.
|
|
|
969
1043
|
- **The transcript** keeps the prompts, every tool call, the agent's text, the
|
|
970
1044
|
commands and file changes Codex reports, and each proposal and what came of it.
|
|
971
1045
|
Each assignment also records what it started on: the toil and Toil UI versions, the
|
|
972
|
-
adapter and model, the tools, and SHA-256 fingerprints of the
|
|
1046
|
+
adapter and model, the tools, and SHA-256 fingerprints of the Toil project, the item and
|
|
973
1047
|
the prompt, and the install fingerprint a later look holds it to (Stale sessions, below). Assignments are kept under the cache, in
|
|
974
1048
|
`<data folder>/cache/assignments/`, so a wipe deletes them.
|
|
975
1049
|
- **For programs on the same server**, `api/assignments` lists, assigns, starts,
|
|
@@ -979,7 +1053,7 @@ you apply it.
|
|
|
979
1053
|
## Stale sessions
|
|
980
1054
|
|
|
981
1055
|
An agent works from the instructions it got when it started. Upgrade toil, edit the
|
|
982
|
-
|
|
1056
|
+
Toil project or change your instructions under **Settings, Agent**, and the chats and
|
|
983
1057
|
assignments already going are still on the old ones. Toil UI records what each one
|
|
984
1058
|
started on, says which have gone stale and why, and renews them on request.
|
|
985
1059
|
|
|
@@ -987,7 +1061,7 @@ started on, says which have gone stale and why, and renews them on request.
|
|
|
987
1061
|
run starts, keeps an install fingerprint: toil's and Toil UI's version and commit,
|
|
988
1062
|
one SHA-256 per toil area of the tools the agent was handed (Toil UI loads no skill
|
|
989
1063
|
page, so those tools are the toil instructions its agent reads), SHA-256 of the
|
|
990
|
-
|
|
1064
|
+
Toil project as `toil projects show` reads it, and SHA-256 of the prompts (the system
|
|
991
1065
|
prompt, the starters or the mode's task, and your own instructions and voice). The
|
|
992
1066
|
theme on screen isn't part of it. An assignment recorded before this has none, and
|
|
993
1067
|
is never called stale.
|
|
@@ -995,7 +1069,7 @@ started on, says which have gone stale and why, and renews them on request.
|
|
|
995
1069
|
assignment page compares what the session started on with the current fingerprint,
|
|
996
1070
|
whose builds are the ones on disk, read again only when their files change. The
|
|
997
1071
|
reason names what moved: "started on toil 0.1.0-alpha.3 · now alpha.4 · skills
|
|
998
|
-
changed: github ·
|
|
1072
|
+
changed: github · Toil project changed". A session's earlier turns aren't marked; only
|
|
999
1073
|
the instructions it started on are.
|
|
1000
1074
|
- **Where it shows.** A **Stale** badge with the reason on the chat (in the Ask bar and
|
|
1001
1075
|
the focused view), on each assignment in the Agents sidebar and on its page, and on the
|
|
@@ -1008,7 +1082,7 @@ started on, says which have gone stale and why, and renews them on request.
|
|
|
1008
1082
|
stays readable, marked as renewed. Renew never interrupts a turn: a chat that is
|
|
1009
1083
|
answering and a working assignment wait, and an assignment that needs you is
|
|
1010
1084
|
skipped until you've reviewed it.
|
|
1011
|
-
- **Renew all stale**, in the Agents sidebar, lists every stale session of the
|
|
1085
|
+
- **Renew all stale**, in the Agents sidebar, lists every stale session of the Toil project
|
|
1012
1086
|
first, the chat included, each saying whether it renews or why it's skipped, and
|
|
1013
1087
|
renews the rest with one press.
|
|
1014
1088
|
- **The server itself.** When toil or Toil UI on disk is newer than the running
|
|
@@ -1042,15 +1116,15 @@ call, its progress, or a state that didn't change never notifies.
|
|
|
1042
1116
|
the views.
|
|
1043
1117
|
- **Where it listens.** The server's `api/assignments/events?since=latest` stream,
|
|
1044
1118
|
which carries what a notice says with each event, and a minute's read of the
|
|
1045
|
-
|
|
1119
|
+
Toil project on screen for sessions gone stale.
|
|
1046
1120
|
|
|
1047
1121
|
### Notification history
|
|
1048
1122
|
|
|
1049
1123
|
The **bell** in the header (beside **Menu** on a narrow screen) counts the unread and
|
|
1050
1124
|
opens the last 50 things worth telling you, newest first: each change a notice carried
|
|
1051
1125
|
(needs you, done, failed, stopped at a usage cap, gone stale), each sync's result, from
|
|
1052
|
-
the Sync panel, with the
|
|
1053
|
-
to what it's about: the assignment, the
|
|
1126
|
+
the Sync panel, with the Toil project that failed, and when each happened. Every entry links
|
|
1127
|
+
to what it's about: the assignment, the Toil project, or the Toil projects view. **Mark all read**
|
|
1054
1128
|
clears the count and keeps the entries. The bell keeps a change that quiet hours held
|
|
1055
1129
|
back, and one whose notice you dismissed; which changes reach it is the notification
|
|
1056
1130
|
settings' choice. It's kept in this browser's `localStorage`, shared by its tabs (with one
|
|
@@ -1062,7 +1136,7 @@ the worklog's facts, never from the bell.
|
|
|
1062
1136
|
Eight built-in themes: Plain, Crate (the default), Harvest, Starship, Realm, Sprawl,
|
|
1063
1137
|
Frontier and Overworld. Pick
|
|
1064
1138
|
one from the header, or under **Settings, General**, where two checkboxes say how far
|
|
1065
|
-
it goes, both
|
|
1139
|
+
it goes, both off by default (#552):
|
|
1066
1140
|
|
|
1067
1141
|
- **Themed wording:** the theme's own lede, eyebrows, dashboard notes, empty
|
|
1068
1142
|
states and loading lines, and its names for a sprint, a board or an epic beside
|
|
@@ -1083,7 +1157,7 @@ reduced-motion setting, which stops everything).
|
|
|
1083
1157
|
|
|
1084
1158
|
Crate, Harvest, Starship, Realm, Sprawl, Frontier and Overworld also have **theme moments**: a launch splash while the first read
|
|
1085
1159
|
runs, their own loading screens with the real progress and a rotating line of their own,
|
|
1086
|
-
an entrance as a
|
|
1160
|
+
an entrance as a Toil project opens, a transition between its tabs, and a short moment when a
|
|
1087
1161
|
sync brings an item done, a build passed or failed, a blocker cleared, a pull request
|
|
1088
1162
|
merged, or a sprint's end. **Theme motion**, under Settings, General, plays them in full,
|
|
1089
1163
|
subtly or not at all, and is off whenever reduced motion is asked for. Every loading
|
|
@@ -1136,22 +1210,22 @@ GitHub CLI signed in (`gh auth login`), or `TOIL_GITHUB_TOKEN` set in the shell.
|
|
|
1136
1210
|
- **The token** comes from `gh auth token` each time it starts and goes only into
|
|
1137
1211
|
Toil UI's environment, as `TOIL_GITHUB_TOKEN`. It's never written to a file or
|
|
1138
1212
|
printed. `GITHUB_TOKEN` and other `TOIL_*` variables in your shell are left out.
|
|
1139
|
-
- **The
|
|
1140
|
-
`toil
|
|
1213
|
+
- **The Toil projects**, made with `toil projects init` and `link` and checked with
|
|
1214
|
+
`toil projects check`: one for `--repo`, and `toil-self` for the public
|
|
1141
1215
|
`imagile-dev/toil` unless `--no-self`. `--repo` defaults to
|
|
1142
1216
|
`imagile-dev/toil-sandbox` (as `sandbox`), which is private to Imagile's
|
|
1143
1217
|
maintainers; anyone else passes their own repository, or seeds a sandbox of their
|
|
1144
1218
|
own (below).
|
|
1145
1219
|
- **toil from this checkout.** Each start builds toil from the clone into
|
|
1146
1220
|
`.scratch/toil` (`npm run build` with `TOIL_BUILD_DIR`, a few seconds), and both the
|
|
1147
|
-
|
|
1221
|
+
Toil project setup and Toil UI run on that build, not on the published toil
|
|
1148
1222
|
`ui/package.json` pins. Toil UI is started with `node --import
|
|
1149
1223
|
ui/scripts/local-toil-hook.mjs`, which resolves `@imagile-dev/toil/api` to
|
|
1150
1224
|
`.scratch/toil/api.js` and says so on stderr. So a command `ui/` calls before its
|
|
1151
1225
|
release works here. The committed `dist/` and `node_modules` are left alone, and a
|
|
1152
1226
|
plain `node ui/dist/server/toil-ui.js` still runs on the pinned toil.
|
|
1153
1227
|
- **Then** it starts `ui/dist/server/toil-ui.js` on loopback and prints its URL.
|
|
1154
|
-
Press **Sync now**, then open a
|
|
1228
|
+
Press **Sync now**, then open a Toil project. `--port <n>` fixes the port, and
|
|
1155
1229
|
`--home <dir>` uses another folder.
|
|
1156
1230
|
- **Ask with Codex.** Install the SDK in the clone without saving it
|
|
1157
1231
|
(`npm install --no-save --ignore-scripts @openai/codex-sdk@0.160.0`), then choose
|
|
@@ -1179,6 +1253,12 @@ GitHub CLI signed in (`gh auth login`), or `TOIL_GITHUB_TOKEN` set in the shell.
|
|
|
1179
1253
|
script says so: run `claude` once in a terminal, which refreshes its own sign-in,
|
|
1180
1254
|
or set a token from `claude setup-token`, and start the script again.
|
|
1181
1255
|
|
|
1256
|
+
To run the same build against your own `~/.toil` instead (your connections and
|
|
1257
|
+
Toil projects as they are, changed only by what you save in Settings, with Toil UI's own
|
|
1258
|
+
data in `.scratch/toil-ui`), run `node ui/scripts/dev-mine.mjs`.
|
|
1259
|
+
That's the way to try unreleased changes on real work (`skills/toil/install.md`,
|
|
1260
|
+
Trying unreleased changes).
|
|
1261
|
+
|
|
1182
1262
|
## Verify before you open a PR: `npm run verify:ui`
|
|
1183
1263
|
|
|
1184
1264
|
One command checks Toil UI the way a reviewer would, and writes what it found to
|
|
@@ -1201,19 +1281,20 @@ npm run verify:ui -- --no-build --starter standup
|
|
|
1201
1281
|
is what the published package runs on.
|
|
1202
1282
|
- **Sample data:** `toil-ui --sample` in a throwaway home, every view through the
|
|
1203
1283
|
layout check (`scripts/layout-check.mjs`) in every theme and scheme at 375, 768 and
|
|
1204
|
-
1440px. On each
|
|
1284
|
+
1440px. On each Toil project's dashboard it also drags a board card to the next column
|
|
1205
1285
|
with real input, a mouse and at 375px a finger, and fails when the card doesn't
|
|
1206
1286
|
follow the pointer or the dropped move doesn't show in the target column. It needs
|
|
1207
1287
|
only a browser, so it runs anywhere, CI included.
|
|
1208
|
-
- **First run:** `toil-ui` on an empty home, with no connections, no
|
|
1209
|
-
`TOIL_*` variable from your shell, through the same check: its
|
|
1288
|
+
- **First run:** `toil-ui` on an empty home, with no connections, no Toil projects and no
|
|
1289
|
+
`TOIL_*` variable from your shell, through the same check: its Toil projects view must
|
|
1210
1290
|
show the three-step checklist, and the palette and the shortcuts sheet open over it.
|
|
1211
1291
|
- **Real data:** the home `dev-real.mjs` sets up, made fresh at
|
|
1212
1292
|
`.scratch/verify/home`, on `imagile-dev/toil-sandbox` with the token from `gh auth
|
|
1213
|
-
token` in Toil UI's environment only. It syncs the `sandbox`
|
|
1214
|
-
same check over the home page, the
|
|
1215
|
-
except that its header may fold into the Menu at 1440px, since a real
|
|
1216
|
-
and links can genuinely fill the row
|
|
1293
|
+
token` in Toil UI's environment only. It syncs the `sandbox` Toil project and runs the
|
|
1294
|
+
same check over the home page, the Toil project and its Board and Dependencies tabs, its worklog and stand-up, one of its items and the settings pages,
|
|
1295
|
+
except that its header may fold into the Menu at 1440px, since a real Toil project's name
|
|
1296
|
+
and links can genuinely fill the row, and its Usage page's failed-command lines are
|
|
1297
|
+
opened on the sample only, since real data may have none.
|
|
1217
1298
|
Without a GitHub sign-in it skips this part and says so; nothing else depends on it.
|
|
1218
1299
|
- **`--ask`** asks Codex one starter question (`hygiene` unless `--starter <id>`) and
|
|
1219
1300
|
checks that an answer arrived, that the server verified every `[[id]]` it cites,
|
|
@@ -1221,7 +1302,7 @@ npm run verify:ui -- --no-build --starter standup
|
|
|
1221
1302
|
run), and that `codex mcp list` in Toil UI's Codex home shows only `toil_ui`. Then it
|
|
1222
1303
|
asks Analyze as the chat's second question, and checks the history of analyses
|
|
1223
1304
|
(#395): that Analyze was kept with its briefing (a recommendation, and its first
|
|
1224
|
-
three reasons as the history's issues); then it changes the
|
|
1305
|
+
three reasons as the history's issues); then it changes the Toil project's WIP limit in the scratch home (`toil projects
|
|
1225
1306
|
wip-limit --replace`), asks Analyze again, checks the history lists both, newest
|
|
1226
1307
|
first, the second compared with the first, and puts the Overview, the Analyses tab
|
|
1227
1308
|
and the second analysis opened through the layout check, shot. It needs the Codex CLI signed in and the
|
|
@@ -1236,7 +1317,7 @@ npm run verify:ui -- --no-build --starter standup
|
|
|
1236
1317
|
- **It stops** every server and browser it started, and exits 1 on any problem.
|
|
1237
1318
|
|
|
1238
1319
|
`--source ado` runs the real-data part against an Azure DevOps Services board
|
|
1239
|
-
instead (the `ado-board`
|
|
1320
|
+
instead (the `ado-board` Toil project, as `dev-real.mjs --source ado` makes it below),
|
|
1240
1321
|
with `TOIL_ADO_BASE_URL`, `TOIL_ADO_PAT` and `--project <name>` (else
|
|
1241
1322
|
`TOIL_ADO_PROJECT`); without them it skips that part and says why. Each source is
|
|
1242
1323
|
one entry in `SOURCES` in `scripts/verify.mjs`. Set `CHROME_PATH` if the browser it finds doesn't draw headless; Playwright's
|
|
@@ -1269,7 +1350,7 @@ toil's `github` area doesn't read.
|
|
|
1269
1350
|
### An Azure DevOps Services board
|
|
1270
1351
|
|
|
1271
1352
|
`--source ado` runs the same way against an Azure DevOps Services project, through
|
|
1272
|
-
toil's `ado` area on a Services connection, with a
|
|
1353
|
+
toil's `ado` area on a Services connection, with a Toil project named `ado-board`:
|
|
1273
1354
|
|
|
1274
1355
|
```bash
|
|
1275
1356
|
node ui/scripts/dev-real.mjs --source ado --project "<project>"
|
|
@@ -1280,7 +1361,7 @@ It needs `TOIL_ADO_BASE_URL` (`https://dev.azure.com/<organization>`) and
|
|
|
1280
1361
|
shell. They go into Toil UI's environment only: toil makes its `ado` (deployment cloud)
|
|
1281
1362
|
connection from them, with writes off, and nothing about the organization is
|
|
1282
1363
|
written to the config file. `--project`, or `TOIL_ADO_PROJECT`, names the
|
|
1283
|
-
project; the
|
|
1364
|
+
project; the Toil project in the scratch home holds it. The Toil project names no repository,
|
|
1284
1365
|
so no pull requests are read; add `repo` to its `ado` section for them.
|
|
1285
1366
|
|
|
1286
1367
|
To fill a board with made-up work, `ui/scripts/ado-seed/` seeds one from Open Work
|
|
@@ -1353,7 +1434,7 @@ npm run screens:sheet -w ui -- --theme realm --width 768 --motion full
|
|
|
1353
1434
|
`ui/.moments/sheets/<theme>-<width>-<motion>.png`, ready to attach to the PR, with its
|
|
1354
1435
|
frames and an HTML copy in a folder beside it. A sheet shows the splash arriving, held and
|
|
1355
1436
|
handing over; the loading screen as its loops turn; each event moment the theme declares
|
|
1356
|
-
coming in, held and going out, cropped to the card; the
|
|
1437
|
+
coming in, held and going out, cropped to the card; the Toil project entrance; and a change of
|
|
1357
1438
|
tab. Every frame is a still: the engine's animations are caught paused as they start and
|
|
1358
1439
|
stepped by hand, so a sheet is the same on every run. `--width` and `--motion` repeat, and
|
|
1359
1440
|
replace the defaults.
|