feat(cli): add Kimi Code support - #452
Merged
Merged
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (52)
🚧 Files skipped from review as they are similar to previous changes (38)
📝 WalkthroughWalkthroughAdds Kimi Code as a supported Trellis platform, including CLI detection and execution, shared/private skill generation, workflow routing, pull-based context handling, platform documentation, and automated coverage. ChangesKimi Code platform integration
Estimated code review effort: 3 (Moderate) | ~25 minutes Sequence Diagram(s)sequenceDiagram
participant User
participant InitCommand
participant KimiConfigurator
participant SkillStorage
participant KimiCLI
User->>InitCommand: init --kimi
InitCommand->>KimiConfigurator: configureKimi(cwd)
KimiConfigurator->>SkillStorage: write shared .agents/skills
KimiConfigurator->>SkillStorage: write .kimi-code/skills
User->>KimiCLI: run Trellis skill
KimiCLI->>SkillStorage: load SKILL.md
KimiCLI-->>User: execute workflow or sub-agent prompt
Possibly related PRs
Suggested reviewers: 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In @.agents/skills/trellis-meta/references/platform-files/platform-map.md:
- Line 50: Update the Kimi Code sub-agent documentation to include plan
alongside coder and explore, keeping all references synchronized. Apply the same
change in
.agents/skills/trellis-meta/references/platform-files/platform-map.md:50-50,
.claude/skills/trellis-meta/references/platform-files/platform-map.md:50-50,
.claude/skills/trellis-meta/references/platform-files/agents.md:37-37, and
.cursor/skills/trellis-meta/references/platform-files/agents.md:37-37.
In `@packages/cli/src/templates/common/bundled-skills/trellis-meta/SKILL.md`:
- Line 15: Remove the stale hard-coded platform count from the bundled source at
packages/cli/src/templates/common/bundled-skills/trellis-meta/SKILL.md:15-15,
updating any nearby references to avoid claiming a fixed number of supported
platforms. Regenerate .agents/skills/trellis-meta/SKILL.md:15-15 from the
corrected bundled source so both copies match.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 7f883ffe-006f-4f1e-84a3-f6325b33b9f5
📒 Files selected for processing (52)
.agents/skills/trellis-meta/SKILL.md.agents/skills/trellis-meta/references/platform-files/agents.md.agents/skills/trellis-meta/references/platform-files/hooks-and-settings.md.agents/skills/trellis-meta/references/platform-files/overview.md.agents/skills/trellis-meta/references/platform-files/platform-map.md.agents/skills/trellis-meta/references/platform-files/skills-and-commands.md.claude/skills/trellis-meta/SKILL.md.claude/skills/trellis-meta/references/platform-files/agents.md.claude/skills/trellis-meta/references/platform-files/hooks-and-settings.md.claude/skills/trellis-meta/references/platform-files/overview.md.claude/skills/trellis-meta/references/platform-files/platform-map.md.claude/skills/trellis-meta/references/platform-files/skills-and-commands.md.cursor/skills/trellis-meta/SKILL.md.cursor/skills/trellis-meta/references/platform-files/agents.md.cursor/skills/trellis-meta/references/platform-files/hooks-and-settings.md.cursor/skills/trellis-meta/references/platform-files/overview.md.cursor/skills/trellis-meta/references/platform-files/platform-map.md.cursor/skills/trellis-meta/references/platform-files/skills-and-commands.md.trellis/scripts/common/active_task.py.trellis/scripts/common/cli_adapter.py.trellis/scripts/common/task_store.py.trellis/spec/cli/backend/platform-integration.md.trellis/workflow.mdREADME.mdREADME_CN.mdmarketplacepackages/cli/src/cli/index.tspackages/cli/src/commands/init.tspackages/cli/src/configurators/index.tspackages/cli/src/configurators/kimi.tspackages/cli/src/templates/common/bundled-skills/trellis-meta/SKILL.mdpackages/cli/src/templates/common/bundled-skills/trellis-meta/references/platform-files/agents.mdpackages/cli/src/templates/common/bundled-skills/trellis-meta/references/platform-files/hooks-and-settings.mdpackages/cli/src/templates/common/bundled-skills/trellis-meta/references/platform-files/overview.mdpackages/cli/src/templates/common/bundled-skills/trellis-meta/references/platform-files/platform-map.mdpackages/cli/src/templates/common/bundled-skills/trellis-meta/references/platform-files/skills-and-commands.mdpackages/cli/src/templates/kimi/agents/trellis-check.mdpackages/cli/src/templates/kimi/agents/trellis-implement.mdpackages/cli/src/templates/kimi/agents/trellis-research.mdpackages/cli/src/templates/kimi/index.tspackages/cli/src/templates/trellis/scripts/common/active_task.pypackages/cli/src/templates/trellis/scripts/common/cli_adapter.pypackages/cli/src/templates/trellis/scripts/common/task_store.pypackages/cli/src/templates/trellis/workflow.mdpackages/cli/src/types/ai-tools.tspackages/cli/test/commands/init.integration.test.tspackages/cli/test/commands/update.integration.test.tspackages/cli/test/configurators/index.test.tspackages/cli/test/configurators/platforms.test.tspackages/cli/test/regression.test.tspackages/cli/test/templates/kimi.test.tspackages/cli/test/templates/trellis.test.ts
Add Kimi Code (kimi) as a first-class platform, modeled on the Pi Agent integration: - Registry: new kimi entry (configDir .kimi-code, supportsAgentSkills, agentCapable, no project-level hooks) and a new /skill:trellis- cmdRefPrefix matching Kimi's /skill:<name> invocation syntax - Configurator: workflow + bundled skills go to the shared .agents/skills/ root via the neutral resolver (byte-identical to Codex/Gemini/Pi writes); session-boundary commands and the trellis-implement/check/research agent prompts ship as Kimi skills under .kimi-code/skills/ (Kimi has no project-level custom sub-agent definitions, only built-in coder/explore/ plan sub-agents) - CLI: --kimi init flag, InitOptions.kimi - Python runtime: cli_adapter.py kimi branches (run: kimi -p --yolo, resume: kimi --session <id>), task_store.py subagent config dirs, active_task.py known platforms; workflow.md marker blocks classify Kimi as a pull-based agent-capable platform - Dogfood copies under .trellis/ kept byte-identical to templates - Tests: templates/kimi.test.ts, configurator detection + byte-parity, init/update integration, regression registry + cli_adapter assertions - Docs: README/README_CN platform lists, platform-integration spec tables, trellis-meta platform-files references - marketplace submodule bumped to the Kimi workflow-mirror sync commit (mindfold-ai/marketplace#10; re-bump to the merge commit after it lands)
lizhangmai
force-pushed
the
feat/kimi-code-support
branch
from
July 21, 2026 05:08
64dfe31 to
0fdc1e9
Compare
…tform count - platform-map.md / agents.md: list all three built-in Kimi sub-agents (coder/explore/plan), matching the platform matrix entry - trellis-meta SKILL.md: remove the hard-coded 'all 15 supported platforms' count (stale; the registry is at 20 and keeps growing) - dogfood copies under .agents/.claude/.cursor kept byte-identical
22 tasks
Xio-Shark
pushed a commit
to Xio-Shark/Trellis
that referenced
this pull request
Sep 17, 2026
* feat(cli): add Kimi Code support Add Kimi Code (kimi) as a first-class platform, modeled on the Pi Agent integration: - Registry: new kimi entry (configDir .kimi-code, supportsAgentSkills, agentCapable, no project-level hooks) and a new /skill:trellis- cmdRefPrefix matching Kimi's /skill:<name> invocation syntax - Configurator: workflow + bundled skills go to the shared .agents/skills/ root via the neutral resolver (byte-identical to Codex/Gemini/Pi writes); session-boundary commands and the trellis-implement/check/research agent prompts ship as Kimi skills under .kimi-code/skills/ (Kimi has no project-level custom sub-agent definitions, only built-in coder/explore/ plan sub-agents) - CLI: --kimi init flag, InitOptions.kimi - Python runtime: cli_adapter.py kimi branches (run: kimi -p --yolo, resume: kimi --session <id>), task_store.py subagent config dirs, active_task.py known platforms; workflow.md marker blocks classify Kimi as a pull-based agent-capable platform - Dogfood copies under .trellis/ kept byte-identical to templates - Tests: templates/kimi.test.ts, configurator detection + byte-parity, init/update integration, regression registry + cli_adapter assertions - Docs: README/README_CN platform lists, platform-integration spec tables, trellis-meta platform-files references - marketplace submodule bumped to the Kimi workflow-mirror sync commit (mindfold-ai/marketplace#10; re-bump to the merge commit after it lands) * docs(cli): address review — sync Kimi sub-agent lists, drop stale platform count - platform-map.md / agents.md: list all three built-in Kimi sub-agents (coder/explore/plan), matching the platform matrix entry - trellis-meta SKILL.md: remove the hard-coded 'all 15 supported platforms' count (stale; the registry is at 20 and keeps growing) - dogfood copies under .agents/.claude/.cursor kept byte-identical
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds Kimi Code (
kimi) as a first-class platform, modeled on the Pi Agent integration (82a44643).Kimi Code reads project-level skills from both
.kimi-code/skills/and the shared.agents/skills/layer (agentskills.io), plusAGENTS.md. It is integrated as a class-2 pull-based platform: agent-capable, but with no project-level hooks (hooks live in user-level~/.kimi-code/config.tomlonly) and no project-level custom sub-agent definitions (only the built-incoder/explore/plansub-agents).Changes
ai-tools.ts): newkimientry —configDir: .kimi-code,supportsAgentSkills: true,agentCapable: true,hasHooks: false; newcmdRefPrefixmember/skill:trellis-matching Kimi's/skill:<name>invocation syntaxconfigurators/kimi.ts):.agents/skills/via the neutral resolver (byte-identical to Codex/Gemini/Pi writes — covered by a dedicated parity test)trellis-start/trellis-continue/trellis-finish-work) and thetrellis-implement/trellis-check/trellis-researchagent prompts →.kimi-code/skills/<name>/SKILL.md(shipped as skills since Kimi has no named sub-agent mechanism; implement/check carry the pull-based prelude; the prompts instruct the main session to dispatch Kimi's built-incoder/exploresub-agent)--kimiinit flag,InitOptions.kimicli_adapter.pykimi branches (run:kimi -p <prompt> --yolo, resume:kimi --session <id>, verified againstkimi --help),task_store.pysubagent config dirs,active_task.pyknown platforms;workflow.mdmarker blocks classify Kimi as pull-based agent-capable.trellis/kept byte-identical to templates (verified withdiff)templates/kimi.test.ts, configurator detection + configure/collect byte-parity, init/update integration, regression registry +cli_adapter+ jsonl-seed assertions, workflow-label mappingplatform-integration.mdspec tables,trellis-metaplatform-files referencesDeliberately out of scope
wire.jsonlsession format needs dedicated adapter work — happy to follow up.Marketplace dependency
marketplacesubmodule pointer to the mirror-sync commit in mindfold-ai/marketplace#10. CI will fail to check out the submodule until that PR lands — please merge the marketplace PR first, then I (or a maintainer) can re-bump the pointer to the merge commit (same flow as the Grok sync in #435).Verification
pnpm test: 1434 passed (63 files), including the marketplace-mirror diff testpnpm typecheck,pnpm lint,pnpm lint:py(basedpyright): cleantrellis init --kimi --yesin a temp repo →.kimi-code/skills/(6 skills),.agents/skills/, AGENTS.md block,/skill:trellis-*refs rendered correctly;trellis update --dry-runreports already-up-to-date (configure/collect hash parity);get_context.py --step 2.1 --platform kimiroutes to the pull-based blockSummary by CodeRabbit
New Features
--kimioption to initialize Kimi Code skills and workflows./skill:trellis-*) and built-in sub-agent workflow integration..agents/skills/and.kimi-code/skills/) without project-level hooks/settings.Documentation