state

GitHub

管理优化任务的完整状态,包括计划文件、测量记录、工作树及临时输出。支持创建、暂停和恢复优化流程,确保各阶段数据一致性与可追溯性。

skills/search/state/SKILL.md uw-syfi/vibesys

Trigger Scenarios

需要创建新的优化任务并初始化状态目录时 中断后需要恢复之前的优化进度时 查询或更新当前优化任务的执行状态与结果时

Install

npx skills add uw-syfi/vibesys --skill state -g -y
More Options

Non-standard path

npx skills add https://github.com/uw-syfi/vibesys/tree/main/skills/search/state -g -y

Use without installing

npx skills use uw-syfi/vibesys@state

指定 Agent (Claude Code)

npx skills add uw-syfi/vibesys --skill state -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": "state",
    "description": "Use when you create, resume or update the state of a vibesys optimization: the plan files, measurement records, worktrees and scratch output under .vibesys-state\/optimize\/<task>\/. Defines the layout, the plan file format and who writes which part, how to find what is live, and how to resume after a stop."
}

Optimization state

One optimization per task, stoppable and resumable at any time. Everything it knows is in .vibesys-state/optimize/<task>/ at the repository root: untracked and listed in .git/info/exclude (the lead creates both on first use). The committed .vibesys/ stays read-only.

plans/         plan-NNN-<slug>.md   one file per idea; NNN zero-padded, ascending, never reused
measurements/  <commit-sha>.jsonl   written ONLY by measure's measurements.py
worktrees/     every git worktree: main, one per plan, judges, throwaway measurement checkouts
scratch/       disposable machine output: logs, raw JSON, profiler output, cached external results

Result branch (the optimization's "main"): vibesys/<task>. Candidates: vibesys-work/<task>/plan-NNN. Pass this directory to measurements.py --dir.

Plan files

One hypothesis per file. The header is three single-line fields, then three sections, each with one writer:

status: proposed|chosen|returned|won|parked|rejected
claim: <one line>
gain: <one line, set by the lead when the plan closes: measured gain with record ids, or why rejected/parked>

## Proposal   (a planner subagent) claim, mechanism, expected gain and its basis, files touched, how to test, evidence
## Decision   (the lead) chosen / rejected / deferred and why; duplicates; disjoint files with other chosen plans
## Result     (the lead, from hand-backs) worker brief text, branch, commit, record ids, judge verdict, lessons

Workers and judges never edit plan files; they hand back evidence. The lead assigns the number. A planner names the file itself once it has chosen its idea.

Cheap index of all plans, without reading bodies:

grep -H -m3 -E '^(status|claim|gain):' plans/*.md
grep -l '^status: \(proposed\|chosen\|returned\)' plans/*.md    # what is live

Measurements

Only measurements.py writes records; never edit a shard. A final (fresh) measurement of a branch is one taken after its last move, on its current head, with several samples. measurements.py check --dir D verifies the result branch head has a passing better record. Slow external jobs (queues, clusters) must be idempotent and cached in scratch/ by commit plus evaluator fingerprint, so a relaunch never repeats or duplicates work.

Resume

  1. Read the open plans (proposed, chosen, returned), then the last few closed ones.
  2. Check git (branches, worktrees, the head of vibesys/<task>) and the running agents for what is real, not what a plan claims.
  3. Relaunch or park every plan whose agent is gone, and say so in its Decision section.
  4. After a long gap or an evaluator change, re-measure main as a control (re-baseline).

Housekeeping

Remove the worktrees of closed plans; keep their branches. Put bulky output in scratch/.

Version History

  • 3a50e8b Current 2026-11-02 13:57

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/chat-skills/steering-a-run/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
f289c73d
Indexed
2026-11-02 13:57

Accueil - Wiki
Copyright © 2011-2026 iteam. Current version is 2.156.0. UTC+08:00, 2026-11-17 23:33
浙ICP备14020137号-1