一键在 Manus 中运行任何 Skill

run

星标0

分支2

更新时间2026年7月15日 07:11

Use to drive a single GitHub issue to a pull request with the kagura-engineer harness — shells out to `kagura-engineer run` with an issue number (guard → recall → worktree → start → implement → ship → persist). HARNESS — this mutates the repo, creates a PR, and spends model budget; confirm with the user before launching.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

kagura-ai

kagura-ai/kagura-engineer

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

kagura-engineer: run

Thin wrapper around the kagura-engineer run CLI verb. It drives one GitHub issue through the full memory-grounded loop to an open PR. It reimplements nothing — the orchestration lives entirely in the CLI; this skill discovers config, gates on doctor, shells out, and surfaces the result.

⚠️ This is a Harness, not an atomic tool. Before launching, make sure the user understands the consequences and has confirmed:

Repo mutation — creates a run-<issue> git worktree and commits there.

PR creation — opens a GitHub pull request via gh-issue-driven on success.

Cost — invokes the selected brain backend multiple times (start / implement / ship).

HITL — gates can halt (blocked, exit 2) and wait for a human decision.

Announce: "Using the kagura-engineer:run skill — this is a Harness that will create a PR."

No-argument usage

If invoked without an issue number, do NOT guess or shell out — print this and stop:

kagura-engineer:run <issue-number>
  Drive one GitHub issue to a PR via the kagura-engineer harness.
  ⚠ Harness: creates a worktree + commits, opens a PR, spends model budget, gates on HITL.
  Example:  kagura-engineer:run 42
  New here? run kagura-engineer:doctor first to check the environment.

Steps

Validate the issue argument. The CLI takes a single positive integer issue number. Reject anything that does not match ^[1-9][0-9]{0,8}$ before interpolating it into the shell — do not pass free-form text through.
Discover config. repo.yaml in CWD by default; pass --config <path> otherwise.
Precondition — run doctor:
```
kagura-engineer doctor --json
```
If overall == "fail" (exit 1), stop: surface the blocking checks and hand off to kagura-engineer:setup. Do not proceed to run. (ok/warn may proceed.)
Confirm with the user (the Harness warning above) unless they have already opted into an unattended run.
Run:
```
kagura-engineer run <issue> --json
```
Add --unattended to proceed without prompting on green/yellow gates (red/unknown still halt), or --no-remember to skip the memory savepoint. Without --json, phase progress streams live.
Interpret the status. status == "ok" (exit 0) → PR created (pr_url). "blocked" (exit 2) → a gate halted; show resume_hint and surface the verdict for a human decision. "fail" (exit 1) → hard error; show the failing phase's detail from the phases array (there is no top-level detail key).
Surface & hand off. Print the phase table and pr_url. On success, extract the PR number from pr_url and suggest kagura-engineer:review <PR-number>. On blocked/fail, surface the resume_hint so the user can retry after addressing the cause.

同仓库更多 Skills

同仓库

setup

kagura-ai/kagura-engineer

Use when kagura-engineer:doctor reports a blocked or failing environment — scaffolds missing config, installs and authenticates harness prerequisites, and generates Memory Cloud MCP config by shelling out to `kagura-engineer setup`, then re-checks with doctor.

2026-07-160

doctor

kagura-ai/kagura-engineer

Use before any kagura-engineer run/review, or when the harness reports a blocked environment — diagnoses the local setup (git, selected brain CLI, gh, ollama, memory backend, and gh-issue-driven plugin) by shelling out to `kagura-engineer doctor` and reports what must be fixed.

2026-07-150

eval

kagura-ai/kagura-engineer

Use to measure whether memory grounding actually improves PR quality — shells out to `kagura-engineer eval` with one or more issue numbers, running the SAME fixed issue set in two arms (recall ON vs OFF) and printing an A/B table. HARNESS — high cost; it runs the full loop twice per issue and mutates the repo. Confirm with the user before launching.

2026-07-150

review

kagura-ai/kagura-engineer

Use to review a git ref, branch, or PR with the local kagura-engineer reviewer — shells out to `kagura-engineer review [target]`, returning a structured verdict and findings. With `--fix` it is a HARNESS that edits files and commits; without `--fix` it is read-only.

2026-07-150

init

kagura-ai/kagura-engineer

Use to bootstrap a fresh checkout for the kagura-engineer harness — scaffolds a commented repo.yaml template and adds it to .gitignore by shelling out to `kagura-engineer init`. Run this before setup/run/review when no repo.yaml exists yet. Idempotent and never overwrites.

2026-06-160

name	run
description	Use to drive a single GitHub issue to a pull request with the kagura-engineer harness — shells out to `kagura-engineer run` with an issue number (guard → recall → worktree → start → implement → ship → persist). HARNESS — this mutates the repo, creates a PR, and spends model budget; confirm with the user before launching.

kagura-engineer: run

⚠️ This is a Harness, not an atomic tool. Before launching, make sure the user understands the consequences and has confirmed:

Repo mutation — creates a run-<issue> git worktree and commits there.

PR creation — opens a GitHub pull request via gh-issue-driven on success.

Cost — invokes the selected brain backend multiple times (start / implement / ship).

HITL — gates can halt (blocked, exit 2) and wait for a human decision.

Announce: "Using the kagura-engineer:run skill — this is a Harness that will create a PR."

No-argument usage

If invoked without an issue number, do NOT guess or shell out — print this and stop:

kagura-engineer:run <issue-number>
  Drive one GitHub issue to a PR via the kagura-engineer harness.
  ⚠ Harness: creates a worktree + commits, opens a PR, spends model budget, gates on HITL.
  Example:  kagura-engineer:run 42
  New here? run kagura-engineer:doctor first to check the environment.

Steps

Validate the issue argument. The CLI takes a single positive integer issue number. Reject anything that does not match ^[1-9][0-9]{0,8}$ before interpolating it into the shell — do not pass free-form text through.
Discover config. repo.yaml in CWD by default; pass --config <path> otherwise.
Precondition — run doctor:
```
kagura-engineer doctor --json
```
If overall == "fail" (exit 1), stop: surface the blocking checks and hand off to kagura-engineer:setup. Do not proceed to run. (ok/warn may proceed.)
Confirm with the user (the Harness warning above) unless they have already opted into an unattended run.
Run:
```
kagura-engineer run <issue> --json
```
Add --unattended to proceed without prompting on green/yellow gates (red/unknown still halt), or --no-remember to skip the memory savepoint. Without --json, phase progress streams live.
Interpret the status. status == "ok" (exit 0) → PR created (pr_url). "blocked" (exit 2) → a gate halted; show resume_hint and surface the verdict for a human decision. "fail" (exit 1) → hard error; show the failing phase's detail from the phases array (there is no top-level detail key).
Surface & hand off. Print the phase table and pr_url. On success, extract the PR number from pr_url and suggest kagura-engineer:review <PR-number>. On blocked/fail, surface the resume_hint so the user can retry after addressing the cause.