feat(kimi-code): recognize multiple inline skill activations in one prompt - #2935
feat(kimi-code): recognize multiple inline skill activations in one prompt#2935chengluyu wants to merge 11 commits into
Conversation
🦋 Changeset detectedLatest commit: 290700d The changes in this PR will be included in the next version bump. This PR includes changesets to release 2 packages
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
|
@codex review |
commit: |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 8e47c73a56
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
8e47c73 to
d98e001
Compare
|
Reworked on top of the refactored engine (#2934 switched from submissionId-correlated groups to bundled prompt messages), force-pushed as a single clean commit. What changed relative to the previous revision:
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: d98e001de3
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Review threads addressed in 000ddca:
The two earlier P2 threads from the pre-bundle revision (replay turn counting across a group, multi-hook-result lookahead) are moot: grouped submissions are a single replay record now, so that machinery was deleted rather than patched. @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 000ddca2a0
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Good catch — the TUI-side preservation was indeed one trim too late: @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: c9bab32228
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Fixed in 17d3d1a: typing token characters inside an inline slash token now counts as an autocomplete context ( @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 17d3d1af52
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Fixed in c3c9467: the retrigger context no longer routes through @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: c3c94670da
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Both fixed in c8c9d4e:
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: c8c9d4ef7c
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Fixed in 82f473a: @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 82f473a6ce
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Semantics decision after design review: skill mentions never carry args in a bundled prompt (42274f9). Two or more skill tokens anywhere in the input — the leading one included — now make one bundled prompt in which every token activates by name only. Args remain a standalone
Docs updated (en/zh): a mention in a prompt never carries arguments. @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 42274f9293
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
Fixed in 0d3c123: the intent is now resolved before the token count is consulted — recognized builtin and plugin commands always keep their own path no matter how many skill tokens their arguments mention, so @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 0d3c1230da
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
…rompt Inline /skill: tokens are recognized anywhere in the prompt (after whitespace, including on following lines) with completion, highlighting, and de-duplication. Submitting goes through session.promptWithSkills, so the engine bundles every activation into the prompt's own user message — one turn, one undo anchor. Replay rebuilds the per-skill cards from the prompt origin's skillActivations and shows only the caller's own parts in the user bubble; undo removes the prompt together with its marked bundle cards; hook results ahead of a bundle are projected inside its window. Enter accepts an inline completion without submitting (pi-tui inlineSlashTrigger), and cache-hint plus /btw pass activations through.
- Mark bundle cards by entry id, not by index into a captured array: the transcript window trim may replace the entries array mid-call. - The replay turn limiter no longer cuts between a bundled prompt and the hook results recorded immediately before it; the oldest visible bundle keeps its hook context. - A leading-combo bundle (/skill:a args /skill:b) now queues while busy like any other inline-skill prompt, instead of being rejected by the single-skill slash gate.
The SDK trims the replay to the requested limit before returning it, so a trim landing between a bundled prompt and its preceding hook results would make them unrecoverable to the TUI-side limiter. Resume now fetches one extra turn of margin; preserveBundleHookResults does the final cut without losing the hook context.
When the terminal delivers `/rev` in one stdin chunk, the slash starts an autocomplete request but the following letters arrived to a null autocomplete state and — unlike a leading slash command — matched no retrigger context, so the stale request was discarded and the menu never appeared. Typing token characters inside an inline slash token now retriggers completion (with a regression test). Also aligns the startup resume tests with REPLAY_FETCH_TURN_LIMIT.
isSlashMenuAllowed confines the slash-command menu to the first line, so reusing it for the inline-slash retrigger context silently disabled retriggering on every later line — the bare-slash request went stale and the menu never appeared. The inline context now covers a token-opening slash on subsequent lines as well (with a regression test).
…ken combos - /btw's initial prompt lives entirely in the slash arguments, so a skill token there sits at position 0; scan it with includeLeading so `/btw /skill:review …` actually activates the skill. - Combo-ness is now decided by the raw inline token count rather than the deduplicated activation count, so `/skill:review check /skill:review` submits as a bundled prompt instead of falling through to the single-skill path with the repeated token swallowed into the args.
A leading combo's first activation carries the raw slash arguments, so a pasted media placeholder in them reached the engine unresolved — unlike the standalone sendSkillActivation path, which rewrites placeholders into escape-proof plain-text file references first. sendInlineSkillUserInput now rewrites any arg-carrying activation the same way (covering the busy queue and /btw intercept paths too), while the media themselves continue to ride the prompt as extracted parts.
Align bundled submissions with the mention model: two or more skill tokens anywhere in the input (the leading one included) make one bundled prompt in which every token activates by name only, and args stay a standalone /skill:<name> args concept. This removes the leading combo's command+args parsing, so the first skill's arguments can no longer leak the next token (displayed as a duplicated prompt under its card), media placeholders no longer need arg rewriting, and newline-separated bundles behave exactly like space-separated ones (parseSlashInput's literal-space separator no longer decides bundle-ness).
…ndle rule The no-args bundle rule claimed any input with two or more skill tokens before checking what led it, so `/btw check /skill:a /skill:b` was submitted to the main agent as a bundled prompt instead of opening the side panel. The intent is now resolved first: builtin and plugin commands always keep their own path regardless of how many skill tokens their arguments mention, while skill-led and newline-led inputs still bundle as before.
…ocal divergences External skill tokens are shaped /skill:<name>, but the inline-slash retrigger character classes excluded ':' — typing the colon launched no replacement request, the bare-slash request went stale, and the menu never appeared for prefixed skill names. Colons now retrigger completion like other token characters (with a regression test). Also registers the inlineSlashTrigger and autocomplete-data divergences in the package's re-vendor protection list.
0d3c123 to
22595ce
Compare
|
Rebased onto the updated base (#2934 was itself rebased onto current main) — merged the queue/steer semantics from main's activate rework with this PR's bundle drain (standalone Also fixed in 22595ce: external skill tokens are shaped @codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 22595ce902
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
…dented inline completion - Ctrl-S steering now stops at the first inline-skill bundle: a later queued message (or the editor draft) no longer jumps ahead of the unsteerable bundle into the running turn, so the conversational order survives steering. - The leading-whitespace slash-path suppression now yields to the inline skill context first, so an indented token (` /skill:rev`) completes like its column-0 equivalent instead of being suppressed as a path.
|
Both fixed in 290700d:
@codex review |
|
Codex Review: Didn't find any major issues. Keep them coming! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
2 participants
Related Issue
Resolve #1736
Stacked on #2934 (engine-side grouped multi-skill submissions). This PR wires the capability up in the TUI.
Problem
The prompt editor only recognizes a single explicit skill activation per prompt. Users who type several
/skill:tokens in one message expect all of them to activate.What changed
TUI (
apps/kimi-code)/skill:tokens are recognized anywhere in the prompt (after whitespace, including on following lines), with completion, highlighting, and de-duplication (utils/inline-skill-tokens.ts).session.promptWithSkills, so the prompt and all activations land as a single grouped turn (one undo unit) using the engine API from feat(agent-core-v2): bundle multiple skill activations into one prompt submission #2934.inlineSlashTrigger+data.inlineSkillmarker)./btw(including the initial prompt) pass activations through.packages/pi-tui: inline slash trigger support in the editor / autocomplete / select-list.Docs updated (
docs/en|zh/guides/interaction.md).Verification
tsc --noEmiton apps/kimi-codepnpm run lint: 0 errorsChecklist
gen-changesetsskill, or this PR needs no changeset.gen-docsskill, or this PR needs no doc update.