Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add xcodethink/open-claude-code-skills --skill 23git clone --depth 1 https://github.com/xcodethink/open-claude-code-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/xcodethink/open-claude-code-skills/23)<a href="https://agentmods.dev/skills/xcodethink/open-claude-code-skills/23"><img src="https://agentmods.dev/badge/skills/xcodethink/open-claude-code-skills/23/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/xcodethink/open-claude-code-skills/23"><img src="https://agentmods.dev/badge/skills/xcodethink/open-claude-code-skills/23.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00413 | $0.07659 |
| Opus 5.5 | $0.00165 | $0.03064 |
| Sonnet 5.5 | $0.00083 | $0.01532 |
| Haiku 4.5 | $0.00041 | $0.00766 |
Grade A, and why
23-跨对话执行纪律 scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 339 lines — stays where its author put it; the contents beside it link to each section on GitHub.
跨对话执行纪律 (Cross-Conversation Execution Discipline) v1.1
解决的真实痛点 (负责人 原话): "AI 联想到新问题就新拓展, 开发一星期核心功能都没实现。" 本质: skill 21 (TEP) 管单个对话不跑偏; 本 skill 管跨多个对话 / 多日 / 多 workstream 的项目不跑偏、不失忆、不重复劳动。 来源: 从
~/Developer/<internal-repo>/incubation/roadmap-discipline/毕业 (ADR D003, 2026-05-31)。MVP 经 8 个自我 dogfood session + 1 个跨项目验证 (ProjectA L0 试点)。
海拔定位 (跟其它元规则 skill 的关系)
现有元规则 skill 按"海拔"分层。本 skill 是跨对话层, 是 19/20/21/22 的 peer, 不是谁的子部分:
| skill | 海拔 | 管什么 |
|---|---|---|
| 19 AI 协作框架 | 行为层 | AI 每对话/每任务怎么说话、怎么主动 |
| 20 机构级项目治理 | 战略层 | 跨 wave / 跨季的 Master Plan / RAID / ADR |
| 21 任务执行纪律 (TEP) | 单对话执行层 | 一个任务/一个对话内 Scope Lock 不跑偏 |
| 22 质量门禁 | 门禁层 | 发版前 Go/No-Go |
| 23 跨对话执行纪律 (本) | 跨对话执行层 | 一个项目跨 N 个对话/多日不跑偏、不失忆 |
关键铁律: 跟 19/20/21/22 互补正交, 冲突时严的赢 (一致 CLAUDE.md 元规则)。 本 skill 不重复 21 已有的单对话 Scope Lock / 硬刹车, 只在它们之上补跨对话维度。
何时触发 (适用边界)
| 项目特征 | 用什么 |
|---|---|
| 单对话能做完 (< 200K context) | 只用 skill 21 TEP |
| 2 对话能做完 | skill 21 + conversation-log |
| ≥ 3 对话 / ≥ 2 周 / ≥ 3 workstream | skill 21 + 本 skill |
| 项目级机构治理 (RAID / ADR / wave) | skill 21 + 本 skill + skill 20 |
机读判定: 项目根 / docs/ / incubation/<sub>/ 任一层级存在 roadmap.yaml → 自动套用本 skill。
缺 roadmap.yaml 但项目明显跨多对话 → 先跟用户对齐是否建立 roadmap (PATH LOCK)。
7 机制总览
┌──────────────────────────────────────────────────────────────────┐
│ 1. PATH LOCK (路径锁) │
│ 多轮讨论 → roadmap.yaml (机读单一事实源) │
│ goal / core_path / sessions[].dod / drift_detection 配置 │
├──────────────────────────────────────────────────────────────────┤
│ 2. SESSION BOOTSTRAP (新对话开局) │
│ AI 必读四件套 → 定位"本对话 = session N" → 第一句话 7 字段 │
│ Scope Card 强制从 roadmap 拉, 不自创 │
├──────────────────────────────────────────────────────────────────┤
│ 3. COCKPIT (进度驾驶舱) │
│ COCKPIT.md = 累计成果 + 进度条 + 当前位置 (roadmap-cli render) │
├──────────────────────────────────────────────────────────────────┤
│ 4. DRIFT WATCH (四层并行漂移检测, roadmap.yaml 内逐层可开关) │
│ L0 reality sync · L1 硬规则 · L2 AI 自检 · L3 人工 │
├──────────────────────────────────────────────────────────────────┤
│ 5. CONVERSATION LOG (项目本地对话日志, 多项目仓防污染) │
│ <project>/conversation-log/YYYY-MM-DD-NN-描述.md (8 段) │
├──────────────────────────────────────────────────────────────────┤
│ 6. L0 REALITY SYNC (plan vs reality 一致性) ★ 本机制 signature │
│ 新对话开局 + 每次 commit 前必跑 R1-R5, 防 plan 跟 git 脱节 │
├──────────────────────────────────────────────────────────────────┤
│ 7. 反虚高/反换轨 加固条款 ★ (v1.1) — 防"自欺" │
│ G1 唯一权威轨道(禁换轨) · G2 验收诚实分级(新建vs verify-only) │
│ G3 反史诗塌缩 · G4 前置门禁识别 · G5 开场补答(跑偏+ParkingLot) │
└──────────────────────────────────────────────────────────────────┘
What ships with it
6 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 339 lines · 413 tokens per session scan A 58798759e616
23-跨对话执行纪律 is a skill published in the GitHub repository xcodethink/open-claude-code-skills (2 stars, last pushed 2mo ago), licensed MIT. It adds 413 tokens to every session and 7,659 once invoked, about $0.0017 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
requirements-writer
Use when converting PM intake into a slice SPEC.md with observable acceptance outcomes, explicit scope, and advisory sibling build dependencies.
batch-all-issues
Resolve every open issue one at a time: fact-check each with web research, close the ones that need no action, and run the goal-pr skill to fix, review, and merge the ones that do — repeating until no actionable issues remain.
post-issue-comment
Post a reply comment on a GitHub issue in English with a natural, concise writing style, based on the intent passed as arguments.
orchestrate
Pipeline orchestration: dispatch the highest-priority ready tasks/work units to agents, manage capacity, and coordinate the Todo to Done flow. Invoked as /agiflow:orchestrate. Uses listtasks, listactivetasksbyorg, listmembers, updatetask, getworkunitprogress.
ticket-craft
Create Jira/Asana/Linear tickets optimized for Claude Code execution - AI-native ticket writing.
brainstorm-feature
SDLC intake that turns a rough idea into an approved BRD-lite brief (Why lane) or delivery contract (Direction lane) at docs/brd/brd-[slug].md, sized by SNC tier.