Sync upstream web console updates into v2.5.0: a Subagents switch that tells the agent to use helpers (per project and skill, sent as a standing instruction only when it changes), a Subagents tab listing the helpers the agent reports, a standalone review that asks what to review instead of assuming the branch diff, a refusal for doc patches that move a version backwards, and the project name in the footer and browser tab title

This commit is contained in:
jparkerweb committed 2026-10-06 17:57:22 -07:00
1 parent 8fb19ec933
commit eac5f6590f
179 files changed
+14955 -466

No files matched your search

@@ -170,7 +170,7 @@ so even a kill at the worst possible moment costs you a repeat, never an answer.
| Exit | Meaning | What you do |
| --- | --- | --- |
| `0` | They pressed Send, or asked a Quick question. stdout is the result JSON, and may carry `workspace` (changes to the folder list, see Workspace). | Handle the cards (if any), the chat (if any) and any `workspace` changes, then post once (below, Quick questions and Workspace). |
| `0` | They pressed Send, or asked a Quick question. stdout is the result JSON, and may carry `workspace` (changes to the folder list, see Workspace) and `subagents` (the helpers switch, see Standing instructions). | Handle the cards (if any), the chat (if any), any `workspace` changes and any `subagents` block, then post once (below, Quick questions, Workspace and Standing instructions). |
| `10` | Still working. stdout has a short progress summary; `pendingWorkspace` counts workspace changes waiting for the next send. | Relay ONE line: the `message` field itself, "Waiting on your answer in the web console", followed by the `terminalLine`. Then call `wait` again. |
| `20` | The server is gone. | `open --resume <sid>`, give them the new link, then keep waiting. |
| `30` | They pressed Cancel. | Stop. Ask what they want to do, in the terminal. |
@@ -647,6 +647,84 @@ for permission is not a failure; report only a read that failed after it.
---
## Standing instructions
On the skills that offer it (`implement`, `implement-review`, `quick-task`,
`review`, `pathfinder`), the person can turn **helpers** (subagents) on or off
from the top bar's Subagents button. Their pick is saved per project and per
skill, off by default, and reaches you as a standing instruction, only when it
changes.
### What arrives
A `subagents` field beside `reply`, and `Subagents:` lines last in `reply`,
after the card lines, the `Quick question:` blocks and the `Workspace:` lines.
It rides on `wait` exit `0`, `chat` exit `0` and `open` output. A setting
change alone never ends a `wait` slice: it waits for the next send, or for
your next `chat` check.
```
Subagents: on (revision 2) — a console setting, not typed by the person.
Hand independent work to helpers if your tools allow it; otherwise carry on alone.
Report each helper on the page (console.md → Standing instructions); questions for the person stay with you.
With no way to start helpers, report one "working alone" entry for this skill run instead.
Hand independent tasks in a phase to helpers: tasks that touch different files and do not depend on each other. Keep tasks that share files, the phase's tests and the sign-off with you.
At most 3 helpers at once.
```
beside `"subagents": { "state": "on", "revision": 2, "max": 3, "instruction": "…" }`.
An edit while on arrives as `Subagents: on (revision 3, replaces revision 2) — …`
with the whole block again; switching off arrives as one line:
`Subagents: off (revision 4) — don't start new helpers for the rest of this session.`
It is a console setting, not something the person typed. It never answers a
card, and it is never recorded under `specs/`.
### Obeying it
- **On:** hand independent work to helpers if your tools allow it, at most
`max` at once, following the instruction lines. `max` is advisory: your
harness's own limit still applies.
- **A higher revision replaces a lower one.** Follow the latest block only.
- **Off:** start no new helpers for the rest of the session. Let running ones
finish and report them.
- **Nothing arrives:** keep doing what the skill says.
- **Scope is the skill run it arrived in.** After a dashboard launch, no block
means the switch is off for this skill, even if an earlier skill in the same
conversation was told "on": `open` starts the count again, and a launch with
the switch off sends nothing.
### Reporting helpers
Post them, at your check-ins, so the page's Subagents tab can show them. The
page never hears from a helper directly:
```jsonc
{ "helpers": [
{ "id": "h1", "title": "Read the store helpers", "ask": "Find every caller of readSubagentSetting.", "state": "running" },
{ "id": "h2", "title": "Check the tests", "ask": "Run the subagents tests and report failures.", "state": "done", "result": "All 9 pass." }
] }
```
- `id` (non-empty), `title` (at most 80 characters), `ask` (at most 400) and
`state` (`running`, `done` or `failed`) are required; `result` (at most 200)
is the one line you add once a helper ends.
- Merged by `id`: post a helper again with its new `state` and `result` and it
updates in place. `"helpers": null` clears the list. A dashboard hop clears it
too; a resume into the same skill keeps it.
- One post per change worth seeing, not a stream.
- Questions for the person stay with you, never with a helper: most harnesses
give helpers no way to ask, and an answer only reaches the agent holding the
session.
- **No way to start helpers?** Post one `{ "id": "alone", "alone": true }`
entry for the skill run instead. The tab then says you are working alone.
- **Pathfinder:** research questions always use helpers and are always
reported, whatever the switch says. The switch governs recon, sketches and
lookups beside the conversation.
- Helpers add no session meter points. Post no `run` for them.
---
## Launches (dashboard only)
A `dashboard` session is not a conversation: the page draws a menu of every
@@ -763,7 +841,10 @@ commands follow the pipeline: the sign-off's is
### Review (`review`)
Scope questions up front, then the findings report as a doc (`id: "report"`),
Scope questions up front — run standalone with nothing named, the first is
a `choice`, "What should I review?", built from a quick scan (branch changes,
uncommitted changes, specs, files you name; each option with its size),
never an assumed target; a build's review skips it (`building.md`). Then the findings report as a doc (`id: "report"`),
then the pick-what-gets-fixed menu as a `multi` or `choice` keyed by finding
id, carrying `"doc": "report"` so the card links straight to the findings.
Settle that item (`"status": "answered"` plus the `answer` they sent) in
@@ -797,6 +878,8 @@ that is what the person is there to download. The finish carries no
- Chart Step 8's research subagents must all report back before the
finish (Finishing → Nothing of yours may still be running); the finish
body then names what each one found.
- Report every research subagent as a helper, whatever the Subagents switch
says (Standing instructions → Reporting helpers).
- At every session end, post `finish` (headline, `body`, and Form A's command as
`command`) BEFORE `stop`. Someone who spent the session in the browser never
sees the Trail Footer, and without this their last screen promises a question
@@ -1146,6 +1229,16 @@ A patch that breaks one of these is rejected with exit `3` and a message:
(Session meter) and a non-empty `id`; a `folderIssue` needs non-empty
`name` and `reason` strings, and `hint`, when present, is one too. Rules 4
and 5 cover `reason` and `hint`.
8. **`helpers` entries are well formed** (Standing instructions → Reporting
helpers): an array of objects, each with a unique `id`, a `title`, an `ask`
and a known `state` within their limits, plus at most one `alone` entry.
Rules 4 and 5 cover `title`, `ask` and `result`.
9. **A doc's `version` never goes back.** Sending a known doc a lower
`version` than the page holds is refused: it is almost always a stale or
copied payload. The same version again is fine, and a dashboard hop starts
every doc fresh. Anything else reusing a doc id in the same session (a
build's second phase, a second report or review) continues from the
version the page holds.
These are workflow rules you already follow. The server just makes them checkable.
@@ -1167,7 +1260,9 @@ cannot see them, and the moment it says the session ended, the person is
free to close that terminal and kill them mid-write. While you wait, post
`{"agent":{"status":"working","activity":"Researching 3 questions (1 of 3 back)"}}`
and update the count as each returns. Never write a `finish.body` that says
work "is running in the background" or "will finish on its own".
work "is running in the background" or "will finish on its own". Mark every
reported helper `done` or `failed` before the finish, so the Subagents tab
never ends on a row still running.
### 1. Post the finish