Kimi cli agent

Skill ImL1s/agent-cli-skills/skills/kimi-cli-agent

Run Moonshot Kimi Code CLI headlessly via kimi-exec.sh. Use when the user asks for kimi, Kimi Code, Moonshot coding agent, or kimi agent profiles (--agent / --agent-file). -r is a soft read-only prompt guard; -p auto-approves tools. On 403 usage limit mark BLOCKED — do not retry-loop. Orchestrator owns git (--no-git).From its SKILL.md

Install

npx -y skills add ImL1s/agent-cli-skills --skill kimi-cli-agent

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

1.8 KB, 448 tokens by cl100k_base, as published. Nobody here has run it

Kimi Code CLI (`kimi`) Executor

When to use

User wants kimi / Kimi Code as the coding agent
Need Kimi agent Markdown profiles or built-in sub-agents from an orchestrator

Wrapper

./kimi-exec.sh -C /path/to/repo -t 1800s "Implement phase A. NO git."
./kimi-exec.sh -r -o json "Review this design; do not edit."
./kimi-exec.sh --agent-file ./agents/reviewer.md -f /tmp/bundle.txt
./kimi-exec.sh -T -C /path/to/repo "Delegate via sub-agents"

Flags: -m (default kimi-code/k3) · -t · -C · -o text|json|stream-json · -f · -r · -T · --agent · --agent-file · --skills-dir · --add-dir · --no-git · -- passthrough.

No PTY needed. Wrapper always runs --output-format stream-json internally and cleans to text/json. -C is absolutized; --add-dir uses the physical workspace path.

Headless permission model

kimi -p auto-approves tools (it is an executor by default)
Interactive --yolo / --auto / --plan cannot be combined with -p
-r is a soft read-only prompt guard — for hard isolation use a throwaway worktree

Native multi-agent

Custom agents as Markdown (--agent / --agent-file); built-in sub-agents; discovery from project/user agent dirs. No dedicated /teamwork slash — use -T + agent profiles.

Quota

On 403 usage limit: do not retry-loop. Mark the seat BLOCKED and continue with other CLIs.

Discipline

Exit code untrusted. Orchestrator owns git. Prefer -f for long bundles.

What ships with it: 4 files

15.8 KB alongside SKILL.md, 4 of them executable

lib/

common.shruns6.8 KB
parse_stream_json.pyruns3.5 KB
pty_run.pyruns1.7 KB

kimi-exec.shruns3.7 KB

Gives 0 of the 12 instructions most context ai engineering skills give in 448 tokens

Counted across 1,328 of the 2,349 authors here whose files we hold, read 2026-09-06

Dispatch a fresh subagent for each taskin 76 of 1328, across 59 files
Perform spec compliance review before code quality reviewin 44 of 1328, across 34 files
Dispatch a final code reviewer after all tasksin 38 of 1328, across 26 files
Answer subagent questions before allowing implementationin 36 of 1328, across 26 files
Use the least powerful model capable of the taskin 33 of 1328, across 26 files
Create a TodoWrite list for all tasksin 32 of 1328, across 22 files
Perform a task review after each implementationin 31 of 1328, across 24 files
Extract all tasks and context from the planin 29 of 1328, across 20 files
Provide full task text to subagentsin 28 of 1328, across 20 files
Use git worktrees for isolated workspacesin 25 of 1328, across 20 files
Specify the model explicitly when dispatching a subagentin 23 of 1328, across 18 files
Execute all tasks from the plan without stoppingin 21 of 1328, across 16 files

Said here and by no other author read

Use -T to prefer subagents
Use --no-git to prevent unintended commits
Use kimi-exec.sh for Kimi Code agent requests
Use -p to auto-approve tools
Use -r for soft read-only prompt guarding
Use -f for long file bundles

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.