Agent Skills › uw-syfi/vibesys › steering-a-run

steering-a-run

GitHub

指导如何向运行中的智能体提议引导指令(Steer),包括设定目标、时机和模式,用于在调用 propose_steer 前或需引导/停止智能体时使用。

resources/chat-skills/steering-a-run/SKILL.md uw-syfi/vibesys

Trigger Scenarios

需要引导或重定向智能体行为时 准备调用 propose_steer 函数前 询问如何停止或调整智能体运行时

Install

npx skills add uw-syfi/vibesys --skill steering-a-run -g -y
More Options

Non-standard path

npx skills add https://github.com/uw-syfi/vibesys/tree/main/resources/chat-skills/steering-a-run -g -y

Use without installing

npx skills use uw-syfi/vibesys@steering-a-run

指定 Agent (Claude Code)

npx skills add uw-syfi/vibesys --skill steering-a-run -a claude-code -g -y

安装 repo 全部 skill

npx skills add uw-syfi/vibesys --all -g -y

预览 repo 内 skill

npx skills add uw-syfi/vibesys --list

SKILL.md

Frontmatter
{
    "name": "steering-a-run",
    "description": "How steers work (target, when, standing or once, withdraw) and how to propose one. Use before calling propose_steer or when asked how to guide, redirect or stop an agent."
}

Steering a run

A steer is guidance in plain words that the run's agents read at their next turn. You propose it with propose_steer(text, target, mode); the operator edits, sends or dismisses it. You cannot send it, and a proposal changes nothing until they send it.

Target says who reads it. Call list_targets first: it shows the exact role, attempt and item ids the run holds now, and propose_steer refuses any other id (the refusal lists the valid ones).

  • {kind: run}: the next turn of any role (the default).
  • {kind: role, id}: that role's turns. See the run-roles skill for what each kind of role does.
  • {kind: attempt, id}: the next turn in that attempt. An attempt that has ended is refused.
  • {kind: item, id}: the next turn of any attempt at that work item (an idea).

Mode says when:

  • next_turn: the target's next turn takes it, once. The right default.
  • standing: every turn of the target from then on, until the operator withdraws it. Use it for a rule that must hold ("keep the weights in bf16"), not for a one-off nudge.
  • interrupt: offered to a turn running now when the provider takes a message mid-turn, otherwise the next turn takes it. Use it only when waiting for the next turn would waste work.

After sending, the run reports whether the steer reached a turn, was delivered mid-turn, or was dropped (its target ended first). A standing steer shows as in force until withdrawn.

Write the text for the agent that reads it: one concrete instruction with its reason, no reference to this chat. A steer cannot change how results are measured. Propose at most one or two steers at a time, say what you chose and why (target and mode), and that the operator decides.

Version History

  • 3a50e8b Current 2026-11-02 13:56

Same Skill Collection

.agents/skills/add-agent-provider/SKILL.md
.agents/skills/add-bundled-skill/SKILL.md
.agents/skills/build-client-feature/SKILL.md
.agents/skills/change-container-images/SKILL.md
.agents/skills/create-issue/SKILL.md
.agents/skills/enforce-rules/SKILL.md
.agents/skills/open-pr/SKILL.md
.agents/skills/review-pr/SKILL.md
.agents/skills/software-design/SKILL.md
.agents/skills/testing/SKILL.md
.agents/skills/vs-init/SKILL.md
.agents/skills/write-agent-prompt/SKILL.md
.agents/skills/write-docs/SKILL.md
.agents/skills/write-skills/SKILL.md
.agents/skills/write-vibesys-tests/SKILL.md
resources/chat-skills/reading-ideas/SKILL.md
resources/chat-skills/results-against-base/SKILL.md
resources/chat-skills/run-roles/SKILL.md
resources/skills/dial9-tokio/skills/dial9-diagnose-long-poll/SKILL.md
resources/skills/dial9-tokio/skills/dial9-red-flags/SKILL.md
resources/skills/dial9-tokio/skills/dial9-runtime/SKILL.md
resources/skills/dial9-tokio/skills/dial9-toolkit/SKILL.md
resources/skills/dial9-tokio/skills/dial9-trace-analysis/SKILL.md
resources/skills/dial9-tokio/skills/dial9-trace-loading/SKILL.md
resources/skills/dial9-tokio/skills/dial9-trace-recipes/SKILL.md
resources/skills/dial9-tokio/skills/dial9-zoom-window/SKILL.md
resources/skills/kernel-correctness/SKILL.md
resources/skills/kernel-ncu-analysis/SKILL.md
resources/skills/kernel-performance/SKILL.md
resources/skills/kernel-rocprof-analysis/SKILL.md
resources/skills/microservice-optimization/SKILL.md
resources/skills/neuron-agentic-development/skills/neuron-nki-debugging/SKILL.md
resources/skills/neuron-agentic-development/skills/neuron-nki-docs/SKILL.md
resources/skills/neuron-agentic-development/skills/neuron-nki-profiling/SKILL.md
resources/skills/serving-systems/SKILL.md
skills/domains/dial9-tokio/dial9-rank-bottlenecks/SKILL.md
skills/domains/dial9-tokio/skills/dial9-diagnose-long-poll/SKILL.md
skills/domains/dial9-tokio/skills/dial9-red-flags/SKILL.md
skills/domains/dial9-tokio/skills/dial9-runtime/SKILL.md
skills/domains/dial9-tokio/skills/dial9-toolkit/SKILL.md
skills/domains/dial9-tokio/skills/dial9-trace-analysis/SKILL.md
skills/domains/dial9-tokio/skills/dial9-trace-loading/SKILL.md
skills/domains/dial9-tokio/skills/dial9-trace-recipes/SKILL.md
skills/domains/dial9-tokio/skills/dial9-zoom-window/SKILL.md
skills/domains/kernel-correctness/SKILL.md
skills/domains/kernel-ncu-analysis/SKILL.md
skills/domains/kernel-performance/SKILL.md
skills/domains/kernel-rocprof-analysis/SKILL.md
skills/domains/kernel-timeline-analysis/SKILL.md

Metadata

Files
0
Version
3a50e8b
Hash
4129534e
Indexed
2026-11-02 13:56

Accueil - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-11-02 18:53
浙ICP备14020137号-1