菜单
Parallel local agent orchestration. Use when running multiple agents on roadmap items simultaneously — to select non-overlapping tasks, set up isolated worktrees, launch agents with proven prompts, monitor progress, handle stalled review cycles, and close out PRs cleanly.
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。
基于 SOC 职业分类
| name | deft-swarm |
| description | Parallel local agent orchestration. Use when running multiple agents on roadmap items simultaneously — to select non-overlapping tasks, set up isolated worktrees, launch agents with proven prompts, monitor progress, handle stalled review cycles, and close out PRs cleanly. |
Read and follow: skills/deft-swarm/SKILL.md
Build a project from a SPECIFICATION.md following Deft framework standards. Use after deft-setup has generated the spec, or when the user has a SPECIFICATION.md ready to implement. Handles scaffolding, implementation, testing, and quality checks phase by phase.
Deterministic structured Q&A interview loop. Use when any skill needs structured user input through focused questions with defaults and confirmation gate.
Iterative pre-PR quality improvement loop. Use before pushing a branch for PR creation -- after completing implementation but before task check. Cycles through Read-Write-Lint-Diff until a full pass produces zero changes.
Greptile bot reviewer response workflow. Use when running a review cycle on a PR — to audit process prerequisites, fetch bot findings, fix all issues in a single batch commit, and exit cleanly when no P1/P2 issues remain. Enables cloud agents to run autonomous PR review cycles.
Structured roadmap refresh workflow. Compares open GitHub issues against ROADMAP.md, triages new issues one-at-a-time with human review, and updates the roadmap with phase placement, analysis comments, and index entries.
Set up a new project with Deft framework standards. Use when the user wants to bootstrap user preferences, configure a project, or generate a project specification. Walks through setup conversationally — no separate CLI needed.
