Workflow & Core Tools · mattpocock/skills

wait-what

Name: wait-what
Author: mattpocock/skills

Stop. That last message did not land: re-pitch it.

L1 Basic · Source registration onlyV1 Commit-PinnedDual-pass Audit

Author mattpocockGitHub mattpocock/skills ↗Stars 253948License MIT（GitHub API spdx 标注；该 commit 的 checkout 树内未发现 LICENSE 文件，git ls-files 亦无）Commit 3cca18b368

Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

The entire skill consists purely of prompt orchestration with no scripts or assets: the body is merely an imperative instruction directing the model to perform a 'communication repair'—acknowledge that the previous message was not understood, supply a small amount of missing context, and re-explain. It relies entirely on the model's linguistic execution of the prompt rather than code logic.

productivity/wait-what/SKILL.md

Wait, I don't understand where you've got to here. Re-pitch that: give me a little bit of context, talk in ASD-STE100 Simplified Technical English, and use the ubiquitous language from `CONTEXT.md` (follow `CONTEXT-MAP.md` to the right one if the repo has more than one).

The skill directory contains only 2 files totaling 552 bytes (SKILL.md 394B + agents/openai.yaml 158B). The body opens from the confused first-person perspective of the user, effectively turning 'I'm stuck' directly into an instruction for the model.

Triggering is explicitly restricted to manual user invocation: `disable-model-invocation: true` on the Claude Code side, and `allow_implicit_invocation: false` on the OpenAI/Codex side, preventing the model from self-triggering; the opening line dictates 'what the model should do when it sees user confusion', rather than allowing the model to proactively call a stop.

productivity/wait-what/SKILL.md

disable-model-invocation: true

Correspondingly annotated as User-invoked in productivity/README.md: 'Reachable only when you type them (Claude Code: `disable-model-invocation: true`; Codex: `policy.allow_implicit_invocation: false` in `agents/openai.yaml`)', corroborating the metadata across both files.

Wording is achieved through dual linguistic constraints: STE100 Simplified Technical English (controlled technical English constraining syntax/vocabulary) compresses expressive complexity; ubiquitous language from CONTEXT.md locks terminology to the repository's unified vocabulary, with multi-context repositories locating the correct CONTEXT.md via CONTEXT-MAP.md first.

productivity/wait-what/SKILL.md

talk in ASD-STE100 Simplified Technical English, and use the ubiquitous language from `CONTEXT.md` (follow `CONTEXT-MAP.md` to the right one if the repo has more than one)

STE100 rules are neither bundled nor linked, relying on the model's built-in knowledge (this skill is the only mention of STE100 in the repository); CONTEXT.md is a local repository file read by the model as instructed.

The CONTEXT.md/CONTEXT-MAP.md convention is not native to wait-what, but rather shared across the mattpocock multi-skill ecosystem (defined natively in engineering/domain-modeling/CONTEXT-FORMAT.md: a single root CONTEXT.md for single-context setups; CONTEXT-MAP.md listing locations and relationships across contexts for multi-context setups; lazily created by skills like domain-modeling). wait-what is a pure consumer.

engineering/domain-modeling/CONTEXT-FORMAT.md

If `CONTEXT-MAP.md` exists, read it to find contexts - If only a root `CONTEXT.md` exists, single context - If neither exists, create a root `CONTEXT.md` lazily when the first term is resolved

The phrase 'follow CONTEXT-MAP.md to the right one' in the body of wait-what reuses this resolution rule, though wait-what itself does not specify fallback behavior when CONTEXT.md is missing.

The author provides its ecosystem positioning in ask-matt: wait-what serves as 'ex-post correction' (remedying messages that did not land, usable mid-session across any skill), complementing the 'ex-ante prevention' of grill-with-docs (agreeing on shared language upfront).

engineering/ask-matt/SKILL.md

`/wait-what` is the corrective for a message that didn't land. Use it mid-conversation, inside any other skill, and the agent re-pitches what it just said with the context you were missing, in plain English, using the `CONTEXT.md` vocabulary. It works after the fact; `/grill-with-docs` is the upfront cure

Sourced from the author's own account in another skill within the repository, consistent with wait-what's own declarations and the description in productivity/README.

2Core Capabilities

01Re-pitch the previous ungrasped message and proactively supply context missing for the user

02Constrain output language to ASD-STE100 Simplified Technical English (controlled technical English)

03Enforce ubiquitous language from the repository's CONTEXT.md; route multi-context repositories to the appropriate CONTEXT.md via CONTEXT-MAP.md

04Explicit user invocation only: model cannot self-trigger (Claude Code metadata)

05Cross-host configuration declaration: provides display_name and short_description for Codex/OpenAI hosts while likewise disallowing implicit invocation

4Recommended Governance & Evidence

Suggested Controls for Safe Operation

For Agent Runtime

Can be auto-loaded

For Administrators

Evidence & Triggers (File Paths & Excerpts)

SKILL.md: Pure prompt instructions and rule text; no script execution or external impact.

Audit Notes: Pure prompt skill: no script execution, no outbound network calls, no credential access, and no local disk writes; all behavior directs the model to rephrase the last message using controlled language (reading CONTEXT.md falls under normal host file reading). The only caveat: it relies on the model's internal knowledge of STE100, meaning enforcement rigor varies by model.

5Second-Pass Verification

[ok] 第一遍断言「无外部 CLI/API/网络/包依赖」 — 两文件逐字节重读：无 URL、无命令、无包引用、无 import；仓库级 grep 显示 STE100 全仓库唯一提及即本 skill，无配套规则文件
[ok] 第一遍断言「目录内无脚本/资产/隐藏文件」 — find 全目录仅 productivity/wait-what/SKILL.md 与 productivity/wait-what/agents/openai.yaml 两文件，无点文件
[ok] disable-model-invocation 语义=仅用户触发 — productivity/README.md 原文「Reachable only when you type them (Claude Code: `disable-model-invocation: true`; Codex: `policy.allow_implicit_invocation: false`)」与两文件元数据一致
[ok] CONTEXT.md/CONTEXT-MAP.md 约定来源（wait-what 自身不携带） — engineering/domain-modeling/CONTEXT-FORMAT.md 定义了单/多 context 布局与读取规则（含「create a root CONTEXT.md lazily」）；setup-matt-pocock-skills/domain.md 也引用该约定；wait-what 目录内确无此类文件
[ok] 作者生态定位描述（ask-matt 提及 wait-what）与正文能力相符性 — ask-matt/SKILL.md 原文「re-pitches what it just said with the context you were missing, in plain English, using the CONTEXT.md vocabulary. It works after the fact」与 SKILL.md 正文指令一一对应，无夸大
[ok] 功能声明 vs 实现相符性 — description「Stop. That last message did not land: re-pitch it.」与正文一致；宣称的只是 prompt 能做的事，无吹牛
[ok] license 元数据 — GitHub API 返回 license.spdx_id=MIT；但本 commit checkout（git ls-files + find）无 LICENSE 文件——记录来源为 API，非树内文件
[ok] stars/last_push 元数据 — API 2026-09-06 采集：stargazers_count=253948，pushed_at=2026-09-04T08:45:43Z 与 HEAD 3cca18b（2026-09-04）一致

6Verdict & Scenarios

Extremely low-friction instant correction: the entire skill is a single body instruction, triggered whenever the user says 'didn't get it' to re-pitch

Terminological discipline: forces a return to the ubiquitous language in CONTEXT.md, avoiding re-explaining while remaining mired in jargon

STE100 controlled language constrains expressive complexity; re-pitching favors short sentences and simple words, beneficial for distracted or non-native speakers

Clear safety boundaries: no scripts, network calls, credentials, or disk writes, and both hosts prohibit model self-invocation, ensuring explicit user-only triggers

Clear ecosystem positioning: the author establishes the division of labor with grill-with-docs in ask-matt (ex-post correction vs. ex-ante prevention), allowing modular combination

Best for: Suitable for: real-time correction when messages repeatedly 'fail to land' during long agent sessions; works best paired with repositories that have an established CONTEXT.md ecosystem (initialized via domain-modeling/setup-matt-pocock-skills); also useful for teaching those unfamiliar with engineering context to interpret agent outputs.

Not for: Not suitable for: standalone task-oriented skills (no deliverables, no workflows, no scripts); scenarios where the repository lacks CONTEXT.md and does not plan to build unified terminology first; situations requiring auditable or compliance-grade STE100 documentation.

Install Agent-Ready

① Verified Mirror 2026-09-06

Option A · Direct DownloadDownload wait-what.tar.gz

sha256: 7953c2143da38542…

Option B · Copy Guide for Agent

install.json

Agents automatically download and verify mirrors from this JSON guide.

② Upstream GitHub

GitHub Repo ↗

Pinned to commit 3cca18b368.