@staff0rd/assist 0.1176.1 → 0.1177.1
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/agent/commands/add-rule.md +1 -1
- package/agent/commands/ask.md +2 -2
- package/agent/commands/bug.md +1 -1
- package/agent/commands/design.md +1 -1
- package/agent/commands/draft.md +1 -1
- package/agent/commands/github.md +1 -1
- package/agent/commands/pr.md +1 -1
- package/agent/commands/refine.md +1 -1
- package/agent/commands/review-config.md +1 -1
- package/agent/commands/review-high-level.md +1 -1
- package/dist/.build-files +103 -101
- package/dist/{addActivity-HMULWVAE.js → addActivity-EVJKYYRK.js} +5 -5
- package/dist/{addPhase-GDFLFVAW.js → addPhase-EMUC75TM.js} +10 -8
- package/dist/{advise-PTDTTAJM.js → advise-3WCBQOWA.js} +10 -8
- package/dist/{ask-2BP3HST6.js → ask-FKL4O7CM.js} +10 -2
- package/dist/{askCheck-MZ5TZQIL.js → askCheck-ASINLULR.js} +14 -12
- package/dist/{askDesign-3GDEDLOS.js → askDesign-3QUHOJPI.js} +9 -7
- package/dist/{askQuestion-DNMDJGPV.js → askQuestion-WYUW7O6I.js} +6 -2
- package/dist/{branch-FEGHBGSM.js → branch-4E2A5OSA.js} +7 -7
- package/dist/built-version +1 -1
- package/dist/{chunk-7IAGWBEG.js → chunk-2AAQSTV7.js} +8 -8
- package/dist/{chunk-3OPLEVRF.js → chunk-2ULGYTC7.js} +1 -1
- package/dist/{chunk-KOHPFCX6.js → chunk-34JCYVZJ.js} +1 -1
- package/dist/{chunk-JENJLKHU.js → chunk-3LZFM4XP.js} +8 -8
- package/dist/{chunk-HO7RNI6V.js → chunk-55DAVOXM.js} +9 -2
- package/dist/{chunk-2YDDDABC.js → chunk-5DC6R45H.js} +1 -1
- package/dist/{chunk-EZRENPTJ.js → chunk-5X376XAZ.js} +2 -2
- package/dist/{chunk-X3HQIMRF.js → chunk-6C2FGUCJ.js} +1 -1
- package/dist/{chunk-SZ3YJ37W.js → chunk-6EGDNDME.js} +4 -4
- package/dist/{chunk-TWHFLOJS.js → chunk-7OHKYQP6.js} +3 -3
- package/dist/{chunk-SE77SM55.js → chunk-AA4KTFRA.js} +1 -1
- package/dist/chunk-AE63ECHM.js +22 -0
- package/dist/{chunk-KYMDHDXC.js → chunk-CCV7FX7F.js} +1 -1
- package/dist/{chunk-MT76WK6E.js → chunk-D2JJTYYP.js} +1 -1
- package/dist/{chunk-MVFHRTPB.js → chunk-EBBTBVRF.js} +2 -2
- package/dist/{chunk-DRWK3DK7.js → chunk-F4EHXLJX.js} +1 -1
- package/dist/{chunk-EWMLTEQ6.js → chunk-GA4GCC2X.js} +1 -1
- package/dist/{chunk-CE3IBI6J.js → chunk-GNDYSVHK.js} +2 -2
- package/dist/{chunk-QUA3MKDI.js → chunk-GPEP64A2.js} +1 -1
- package/dist/{chunk-FZZRFRDB.js → chunk-HMQIAEJG.js} +3 -3
- package/dist/{chunk-R7PLAKNF.js → chunk-I6OVO66A.js} +2 -2
- package/dist/{chunk-SGYCDSBF.js → chunk-IVSNDH4D.js} +2 -2
- package/dist/{chunk-MSWG2GGI.js → chunk-KQLM2B42.js} +1 -1
- package/dist/{chunk-QWFK7Y2O.js → chunk-LDNUAKUV.js} +1 -1
- package/dist/{chunk-PXQTFJ2N.js → chunk-LKZ7J5AK.js} +1 -1
- package/dist/{chunk-ICVMDDIK.js → chunk-LYNHUQIU.js} +7 -7
- package/dist/{chunk-MTN37C4V.js → chunk-N4ZFTKU5.js} +1 -1
- package/dist/{chunk-7DTTXLS3.js → chunk-NISL2DNR.js} +1 -1
- package/dist/{chunk-7WQ6XW6H.js → chunk-NTHIX33Z.js} +1 -1
- package/dist/{chunk-SENG2BJD.js → chunk-NX4Q3JM3.js} +1 -1
- package/dist/{chunk-7HOOPQCG.js → chunk-OTY4ZZK5.js} +2 -2
- package/dist/{chunk-RHE5G4R7.js → chunk-S35U37KB.js} +3 -3
- package/dist/{chunk-PA3Z7HIH.js → chunk-S43ZXLSG.js} +1 -1
- package/dist/{chunk-V2UWYFAM.js → chunk-SEYMNOHZ.js} +7 -7
- package/dist/{chunk-AM4HAVX5.js → chunk-SVOKDZOH.js} +1 -1
- package/dist/{chunk-GY2735TY.js → chunk-TFNKFLFS.js} +4 -4
- package/dist/{chunk-J6IDOIVY.js → chunk-TTGFYYM3.js} +1 -1
- package/dist/{chunk-HU2MRZOA.js → chunk-VDXZYKAV.js} +1 -1
- package/dist/{chunk-EWLY7COZ.js → chunk-VJSCW57R.js} +4 -4
- package/dist/{chunk-WXVVO4LF.js → chunk-VO7AQC4M.js} +1 -1
- package/dist/{chunk-YPPA5Z3U.js → chunk-W3WCK6CC.js} +8 -8
- package/dist/{chunk-4QU25WPX.js → chunk-W5YG7STH.js} +1 -1
- package/dist/{chunk-RGMAUOOR.js → chunk-WNHEHLLU.js} +1 -1
- package/dist/chunk-X7PGOOAB.js +26 -0
- package/dist/{chunk-YOE2FWGH.js → chunk-XHPQVAVX.js} +2 -0
- package/dist/chunk-XJTXAL22.js +8 -0
- package/dist/{chunk-RILCJ3PN.js → chunk-XKEY2S2G.js} +14 -14
- package/dist/{cliHook-7XUKUZMI.js → cliHook-MX7SGXI5.js} +11 -8
- package/dist/{cliHookCheck-IBH5EDWF.js → cliHookCheck-LEPJJOB2.js} +4 -4
- package/dist/{codexHook-XZCX2IML.js → codexHook-FPWKVVXR.js} +14 -12
- package/dist/commands/sessions/web/bundle.js +81 -81
- package/dist/{comment-OGILDYP4.js → comment-T6VKMQSC.js} +11 -3
- package/dist/{commentIssue-NSVPWOO4.js → commentIssue-KYYGEMM6.js} +10 -2
- package/dist/{commit-DVBBSH6B.js → commit-BQI7L6WM.js} +5 -5
- package/dist/{createIssue-EINOQRWO.js → createIssue-RDI2HOU4.js} +9 -1
- package/dist/{doctorNodes-CXISKNGM.js → doctorNodes-KLJGQQ76.js} +3 -3
- package/dist/{done-3UYUB6T4.js → done-JJOEGPGX.js} +8 -8
- package/dist/{drainDaemon-PGCYYZVB.js → drainDaemon-LRMJS4UO.js} +1 -1
- package/dist/{edit-76E3EHPS.js → edit-SDMU64FM.js} +9 -1
- package/dist/{editHook-DWCUXU4G.js → editHook-UXIHATIT.js} +11 -8
- package/dist/{editIssue-75CKKMHB.js → editIssue-2BVQHWJT.js} +9 -1
- package/dist/{editIssueComment-DNTA3DUX.js → editIssueComment-RORO62JH.js} +10 -2
- package/dist/{fixConflict-6SCGNOHF.js → fixConflict-GE72P7WJ.js} +7 -7
- package/dist/{harnessFollowUps-6SISMYAK.js → harnessFollowUps-6XB5AOCC.js} +1 -1
- package/dist/index.js +64 -64
- package/dist/{launchMode-MLGYLVZG.js → launchMode-R3QYAUZ7.js} +17 -17
- package/dist/{list-TP35UEF4.js → list-4J7IDSCX.js} +9 -9
- package/dist/{next-OISQZZ4J.js → next-2R5FZLHC.js} +16 -16
- package/dist/{piHook-RGPHKBJR.js → piHook-TWZ7VL4C.js} +11 -8
- package/dist/{postSlackMessage-KO3TWHCM.js → postSlackMessage-W4PBW5GV.js} +10 -2
- package/dist/{propose-P4FKRBJG.js → propose-ARHZAMP3.js} +11 -9
- package/dist/{raise-AM7URDOG.js → raise-5RO44AHM.js} +9 -1
- package/dist/{refine-Q2KCM2VH.js → refine-PHFWOBXG.js} +18 -18
- package/dist/{reply-UKSLXFXX.js → reply-4NHJUAW2.js} +10 -2
- package/dist/{restartNode-ZRORHTS4.js → restartNode-QKP2PI3P.js} +5 -5
- package/dist/{review-BSVYVLJU.js → review-24PY4QIO.js} +12 -10
- package/dist/{reviewPrComments-TNFVZID4.js → reviewPrComments-DI55I3RN.js} +7 -7
- package/dist/{rewindPhase-D2W6JK2M.js → rewindPhase-PYPWUAQM.js} +12 -12
- package/dist/{rulesAdd-BUCJ52JM.js → rulesAdd-275CED56.js} +10 -2
- package/dist/{run-Y42G36KG.js → run-A6U6NHHX.js} +14 -14
- package/dist/{run-WMDSFFZX.js → run-AD4YXUMT.js} +14 -14
- package/dist/{runDaemon-OUVMWOZ4.js → runDaemon-XHPC4BU4.js} +161 -55
- package/dist/{runExtract-FKTXZBLR.js → runExtract-O2VT6GXC.js} +9 -1
- package/dist/{runRefine-SO4LXPEG.js → runRefine-LHELOTJM.js} +18 -18
- package/dist/{scan-WHCF7SGN.js → scan-2DJUBCO7.js} +6 -6
- package/dist/{schemaConfirm-F3KHOXQK.js → schemaConfirm-YQ75EKIB.js} +4 -4
- package/dist/{schemaUnlock-XZZXKAYM.js → schemaUnlock-7RER37EP.js} +4 -4
- package/dist/{search-XSLH3HO4.js → search-4REYRJSL.js} +9 -9
- package/dist/{show-CISZ7PE7.js → show-ASBGNNIT.js} +6 -6
- package/dist/{show-Z4ECMAUB.js → show-YVFFZ342.js} +7 -7
- package/dist/{signExtension-UZQEIYK6.js → signExtension-A6JWCGRO.js} +1 -1
- package/dist/{slack-UG5YUS2B.js → slack-I2CXIEBY.js} +10 -2
- package/dist/{updateNode-6SGAI3J2.js → updateNode-5GRMK3HC.js} +5 -5
- package/dist/{updatePlan-YEAFHGHN.js → updatePlan-5ZHUWXQO.js} +11 -9
- package/dist/{web-FTLFSZKP.js → web-3LF3Y3ZN.js} +20 -20
- package/dist/{web-BENPC2SK.js → web-ZVEWYMUV.js} +20 -20
- package/dist/{wontfix-RKRTM5HV.js → wontfix-LJSLOGTK.js} +10 -2
- package/package.json +1 -1
- package/dist/chunk-TQ5KKZ6G.js +0 -8
|
@@ -59,7 +59,7 @@ It appends the rule to `advice.rules` in `user (db)` with no R-code. For `user (
|
|
|
59
59
|
|
|
60
60
|
Use single quotes around the rule text to avoid shell escaping issues.
|
|
61
61
|
|
|
62
|
-
Run `assist rules add` as a background task and do no other work until it returns. In a web session it opens the rule, its title and where it will land in the preview pane, and writes nothing until the user decides. In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run `assist rules add` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
62
|
+
Run `assist rules add` as a background task and do no other work until it returns. In a web session it opens the rule, its title and where it will land in the preview pane, and writes nothing until the user decides. In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run `assist rules add` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
63
63
|
|
|
64
64
|
- **Approved** — it exits 0 having written the rule. Show the command output — it names the allocated code (or, for an `advice.rules` rule, the title) and where it went — then tell the user in one line which rule was added and where.
|
|
65
65
|
- **Rejected** — it exits non-zero having written nothing, printing the reason and every inline comment. Revise the wording, title or scope to address each one, then re-run `assist rules add` with the revision again, the same way. Do not move on to Step 5 until a version is approved.
|
package/agent/commands/ask.md
CHANGED
|
@@ -21,14 +21,14 @@ assist ask --title '<short title>' --body - < <scratch file>
|
|
|
21
21
|
|
|
22
22
|
How to run it depends on the session:
|
|
23
23
|
|
|
24
|
-
- **Claude SDK session** (`ASSIST_SDK_AGENT=1` in the environment) — run it in the foreground. It opens the preview and returns at once, printing that the preview is open. End your turn and do nothing else: when the user decides, the command's output and exit code arrive as your next message. To revise the update while it is still pending, re-run Step 2; the earlier preview is superseded and delivers nothing.
|
|
24
|
+
- **Claude SDK session** (`ASSIST_SDK_AGENT=1` in the environment) **or Codex session hosted by the assist daemon** (`ASSIST_SESSION=1`) — run it in the foreground. It opens the preview and returns at once, printing that the preview is open. End your turn and do nothing else: when the user decides, the command's output and exit code arrive as your next message. To revise the update while it is still pending, re-run Step 2; the earlier preview is superseded and delivers nothing.
|
|
25
25
|
- **Any other session** — run it **as a background task**, and do no other work until it returns. In a web session it blocks until the user decides, which can take far longer than the default command timeout, and the pending preview dies with the process. To revise the update while it is still pending, re-run Step 2 directly — do not stop the earlier task. The pane swaps to the new version, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.`: ignore it.
|
|
26
26
|
|
|
27
27
|
Outside a web session it prints the markdown and exits 0 without a decision: ask the user in chat instead.
|
|
28
28
|
|
|
29
29
|
## Step 3: act on the decision
|
|
30
30
|
|
|
31
|
-
The decision is the command's output: the background task's result, or in a Claude SDK session the next message.
|
|
31
|
+
The decision is the command's output: the background task's result, or in a Claude SDK or daemon-hosted Codex session the next message.
|
|
32
32
|
|
|
33
33
|
- **Approved (exit 0)** — any inline comments are printed, each with its quoted excerpt. Proceed with the update, folding in every comment.
|
|
34
34
|
- **Rejected (non-zero exit)** — the reason and every inline comment are printed. Do not proceed. Address each one, revise the scratch file, and re-run Step 2.
|
package/agent/commands/bug.md
CHANGED
|
@@ -91,7 +91,7 @@ Omit `phases` when you are not authoring a plan. Each phase needs at least one t
|
|
|
91
91
|
|
|
92
92
|
The payload is strict JSON — an unknown key is an error. `\n` inside the `description` string is a JSON escape and becomes a real newline, which is what the markdown rendering needs.
|
|
93
93
|
|
|
94
|
-
Always run `propose` **as a background task**, and do no other work until it returns. In a web session it blocks on the preview pane until the user decides, which can take far longer than the default command timeout, and the pending preview dies with the process — a killed `propose` abandons the item. To revise the report while its preview is still pending, re-run `propose` with the revised payload directly — do not stop the earlier task. The pane swaps to the new version, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without creating anything: ignore it. In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run `propose` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
94
|
+
Always run `propose` **as a background task**, and do no other work until it returns. In a web session it blocks on the preview pane until the user decides, which can take far longer than the default command timeout, and the pending preview dies with the process — a killed `propose` abandons the item. To revise the report while its preview is still pending, re-run `propose` with the revised payload directly — do not stop the earlier task. The pane swaps to the new version, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without creating anything: ignore it. In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run `propose` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
95
95
|
|
|
96
96
|
`propose` decides where the bug report is reviewed and prints what to do next. Follow that instruction; never inspect the environment to work it out yourself:
|
|
97
97
|
|
package/agent/commands/design.md
CHANGED
|
@@ -50,7 +50,7 @@ assist ask-design --url '<artifact-url>' --title '<design title>' --body - < <su
|
|
|
50
50
|
- **Approved** (exit 0) — fold every printed comment into the design, republishing to the same URL, then move on.
|
|
51
51
|
- **Rejected** (non-zero, printing the reason and comments) — revise the design to address each one, republish to the same artifact URL, and re-run `ask-design` for the next round. Repeat until approved.
|
|
52
52
|
- To revise while a round is still pending, re-run `ask-design` directly; the earlier task exits 0 as superseded — ignore its result.
|
|
53
|
-
- In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run `ask-design` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
53
|
+
- In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run `ask-design` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
54
54
|
- If it prints the design link and exits 0 without `Approved.` (not in an assist web session), ask the user in chat instead and iterate there — editing the same file and republishing to the same URL — until they say they are happy with it.
|
|
55
55
|
- Without an `Artifact` tool there is no artifact URL, and `ask-design` refuses with the reason design review is unavailable on this harness. Review in chat instead: give the user the HTML file's absolute path and iterate on the same file until they say they are happy with it, then skip the link and attach steps below and summarise the design in the item's description.
|
|
56
56
|
|
package/agent/commands/draft.md
CHANGED
|
@@ -121,7 +121,7 @@ JSON
|
|
|
121
121
|
|
|
122
122
|
The payload is strict JSON — an unknown key is an error. `\n` inside the `description` string is a JSON escape and becomes a real newline, which is what the markdown rendering needs. Each phase needs at least one task; omit `manualChecks` for the phases that don't need any (most of them). A phase named `Review` is rejected: the runner appends that phase itself.
|
|
123
123
|
|
|
124
|
-
Always run `propose` **as a background task**, and do no other work until it returns. In a web session it blocks on the preview pane until the user decides, which can take far longer than the default command timeout, and the pending preview dies with the process — a killed `propose` abandons the item. To revise the item while its preview is still pending, re-run `propose` with the revised payload directly — do not stop the earlier task. The pane swaps to the new version, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without creating anything: ignore it. In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run `propose` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
124
|
+
Always run `propose` **as a background task**, and do no other work until it returns. In a web session it blocks on the preview pane until the user decides, which can take far longer than the default command timeout, and the pending preview dies with the process — a killed `propose` abandons the item. To revise the item while its preview is still pending, re-run `propose` with the revised payload directly — do not stop the earlier task. The pane swaps to the new version, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without creating anything: ignore it. In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run `propose` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
125
125
|
|
|
126
126
|
`propose` decides where the item is reviewed and prints what to do next. Follow that instruction; never inspect the environment to work it out yourself:
|
|
127
127
|
|
package/agent/commands/github.md
CHANGED
|
@@ -64,7 +64,7 @@ Display the output so the user can see what happened. If the command exits non-z
|
|
|
64
64
|
|
|
65
65
|
Any preview in this skill (`edit`, `create`, `comment`) can be revised while still pending by re-running the command directly — do not stop the earlier task. The earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without touching GitHub: ignore it.
|
|
66
66
|
|
|
67
|
-
In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run any of these previewing commands in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
67
|
+
In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run any of these previewing commands in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
68
68
|
|
|
69
69
|
## Action: create
|
|
70
70
|
|
package/agent/commands/pr.md
CHANGED
|
@@ -6,7 +6,7 @@ Run `assist prs raise --help` and follow its guidance exactly to compose and rai
|
|
|
6
6
|
|
|
7
7
|
## After a successful raise
|
|
8
8
|
|
|
9
|
-
In a Claude SDK session (`ASSIST_SDK_AGENT=1`) a web-previewed raise returns at once: end your turn and do nothing else, and its outcome arrives as your next message.
|
|
9
|
+
In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`) a web-previewed raise returns at once: end your turn and do nothing else, and its outcome arrives as your next message.
|
|
10
10
|
|
|
11
11
|
Only once the PR has actually been placed (skip this if the raise was rejected in the web preview or errored — there is nothing to watch), start watching CI in the background so you can keep working:
|
|
12
12
|
|
package/agent/commands/refine.md
CHANGED
|
@@ -64,7 +64,7 @@ The payload is the whole plan, not a patch: every phase you want the item to end
|
|
|
64
64
|
|
|
65
65
|
Do not chain `add-phase`, `update-phase`, `remove-phase` and `move-phase` calls to build up a change — that is one preview per command and leaves the item half-edited if the user rejects part way. Those commands remain for a human typing a quick one-liner at a terminal.
|
|
66
66
|
|
|
67
|
-
In a web session this renders the whole change into the preview pane as one diff — each phase marked unchanged, added, edited or moved, removed phases listed separately, phases the item has already completed flagged, and any clamp of `currentPhase` into the shorter plan shown at the top — and blocks until the user decides. The review can take far longer than the default command timeout, so run `update-plan` **as a background task** and do no other work until it returns; the pending preview dies with the process. To revise the change while it is still pending, re-run `update-plan` with the revised payload directly — do not stop the earlier task. The pane swaps to the new diff, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without writing anything: ignore it. In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run `update-plan` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
67
|
+
In a web session this renders the whole change into the preview pane as one diff — each phase marked unchanged, added, edited or moved, removed phases listed separately, phases the item has already completed flagged, and any clamp of `currentPhase` into the shorter plan shown at the top — and blocks until the user decides. The review can take far longer than the default command timeout, so run `update-plan` **as a background task** and do no other work until it returns; the pending preview dies with the process. To revise the change while it is still pending, re-run `update-plan` with the revised payload directly — do not stop the earlier task. The pane swaps to the new diff, and the earlier task exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` without writing anything: ignore it. In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run `update-plan` in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
68
68
|
|
|
69
69
|
When it returns:
|
|
70
70
|
|
|
@@ -48,7 +48,7 @@ assist ask --title 'High-level review config' --body - < <scratch file>
|
|
|
48
48
|
- **Rejected (non-zero exit)** — write nothing. Address the reason and each comment, re-check any changed glob still matches files, revise the scratch file and re-run `assist ask`.
|
|
49
49
|
- **Superseded (exit 0)** — you re-ran `assist ask` while this one was pending, which is how to revise it; do not stop the earlier task first. Ignore this result and wait for the newer one.
|
|
50
50
|
|
|
51
|
-
In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run it in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
51
|
+
In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run it in the foreground instead: it returns at once, so end your turn and do nothing else — the command's output and exit code arrive as your next message, and a superseded run delivers nothing.
|
|
52
52
|
|
|
53
53
|
Outside a web session `assist ask` only prints the markdown and exits 0 without a decision: ask the user in chat instead, and write only once they have answered.
|
|
54
54
|
|
|
@@ -11,7 +11,7 @@ assist review --high-level $ARGUMENTS
|
|
|
11
11
|
|
|
12
12
|
If this is a re-invocation in a session that already ran it, re-run it anyway — never assume the checklist is still open, since the pane may have gone even while a previous task still looks alive. Do not stop a still-running previous `assist review --high-level` task: launch it afresh, and the new checklist replaces the old one in place. The earlier task then exits 0 printing `Superseded by a newer preview — no decision; ignore this result.` — ignore it.
|
|
13
13
|
|
|
14
|
-
Run it as a background task. In an assist web session it opens the checklist in the web UI preview pane and returns once the user approves or requests changes; otherwise it prints the checklist and exits straight away. Do not review the PR yourself or post anything to GitHub. In a Claude SDK session (`ASSIST_SDK_AGENT=1`), run it in the foreground instead: it returns at once, and the review's output and exit code arrive as your next message once it closes; a superseded run delivers nothing.
|
|
14
|
+
Run it as a background task. In an assist web session it opens the checklist in the web UI preview pane and returns once the user approves or requests changes; otherwise it prints the checklist and exits straight away. Do not review the PR yourself or post anything to GitHub. In a Claude SDK session (`ASSIST_SDK_AGENT=1`) or a Codex session hosted by the assist daemon (`ASSIST_SESSION=1`), run it in the foreground instead: it returns at once, and the review's output and exit code arrive as your next message once it closes; a superseded run delivers nothing.
|
|
15
15
|
|
|
16
16
|
Once it is launched, say in one line that the high-level review has started and end your turn straight away. Do not say a pane is open — you only learn which case applies from the task's output. Do not wait on the task — no polling its output, `sleep`, `Monitor` or reading its output file — and the "do no other work until it returns" rule for sign-off commands does not apply here. Input the user sends while you are mid-turn only queues behind it, so staying in the turn blocks every question until the review closes.
|
|
17
17
|
|