Typescript Ai Native Sweep logo

Typescript Ai Native Sweep

Organization
vibevm
typescript-ai-native-sweep

Run the recurring AI-Native discipline sweep on this TypeScript project — the seven-step floor first, then the health collector's ratchet items. Use daily or several times a day on an active tree; any single item is a safe stop.

Overview

Publishervibevm
Repositoryvibevm
Skill nametypescript-ai-native-sweep
Stars
285
Forks
83
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by vibevm on GitHub. Read the source before you install it.

Installation

Install the Typescript Ai Native Sweep AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/vibevm/vibevm/tree/main/vibevm/vibepacks/org.vibevm.ai-native/typescript-ai-native-lang/v1.0.0/vibevm/vibespecs/skills/typescript-ai-native-sweep
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/vibevm/vibevm.git /tmp/vibevm
mkdir -p .claude/skills
cp -r /tmp/vibevm/vibevm/vibepacks/org.vibevm.ai-native/typescript-ai-native-lang/v1.0.0/vibevm/vibespecs/skills/typescript-ai-native-sweep .claude/skills/typescript-ai-native-sweep
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Typescript Ai Native Sweep in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Typescript Ai Native Sweep on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Typescript Ai Native Sweep is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

The discipline sweep (TypeScript stack) {#root}

@fact:RUNNING-THE-STANDING-SWEEP You are running the standing sweep from the Discipline's Sweep Playbook (spec://org.vibevm.ai-native/core-ai-native/04-SWEEP-PLAYBOOK — the shipped copy is at vibedeps/flow-core-ai-native/<version>/spec/04-SWEEP-PLAYBOOK.md; read it once per session if you have not). @status:impl/done

@fact:two-truths-lead The two truths: @status:impl/done

  • @fact:TRUTH-GATES-ARE-THE-FLOOR the gates are the floor, the sweep is the ceiling, @status:impl/done
  • @fact:TRUTH-GATE-IS-TRUTH and the gate is truth, the collector is a guide. @status:impl/done

@fact:NEVER-SWEEP-ON-A-RED-TREE Never sweep on a red tree. @status:impl/done

@fact:ACT-ON-COLLECTOR-FACTS Act on collector facts, never on memory. @status:impl/done

@fact:ALL-COMMANDS-ARE-THE-SHIPPED-TOOLCHAIN All commands below are the shipped toolchain. @status:impl/done

@fact:IF-NOT-ON-PATH-INSTALL-OR-RUN-IN-PLACE If typescript-ai-native is not on PATH, either install it once — cargo install --path vibedeps/<stack-slot>/crates/typescript-ai-native-cli — or run it in place: cargo run --manifest-path vibedeps/<stack-slot>/Cargo.toml -p typescript-ai-native-cli --bin typescript-ai-native -- <args>. @status:impl/done

Tier 0 — the hard floor (ALWAYS first) {#tier-zero}

sh
typescript-ai-native floor

@fact:FLOOR-HAS-SEVEN-STEPS Seven steps: prettier → tsc → tests → eslint → conform → specmap → test-gate. @status:impl/done

@fact:RED-FLOOR-ADMITS-ONLY-GREENING-WORK Red? The only legal work is making it green — fix, do not proceed. @status:impl/done

@fact:CHECK-THE-PRINTED-POLICY-LINES Check the printed policy lines: a NO conform.toml — topology default in force line means the project is not bootstrapped (typescript-ai-native init) and any green under it is vacuous, and every DISABLED by policy line is a standing decision to re-question weekly — a floor that shrank quietly is the failure mode this line exists to catch. @status:impl/done

Tier 1 — the ratchet (every run) {#tier-one}

sh
typescript-ai-native health

@fact:READ-THE-HEALTH-SUMMARY Read the summary (the JSON at discipline/health/latest-typescript.json is the work-list; its git diff is the trend). @status:impl/done

@fact:take-cheapest-wins-lead Take one or two cheapest wins: @status:impl/done

  1. @fact:RATCHET-DANGER-BAND-FILES danger-band files — split any file at the top of the [540,600) band before an edit trips the 600 budget; the new module keeps (or gains) its own @scope marker so the orphan gate never regresses. @status:impl/done
  2. @fact:RATCHET-UNREASONED-SUPPRESSIONS unreasoned suppressions — every @ts-expect-error WITHOUT -- reason in the census is unrecorded testimony: add the reason or fix the underlying type. @ts-ignore is never acceptable — replace with @ts-expect-error -- reason and watch it fail when the error goes. @status:impl/done
  3. @fact:RATCHET-EXPORT-DOC-EXAMPLE-COVERAGE export doc-example coverage — exports without an @example (or fenced block) are retrieval gaps; document the highest-traffic seam first. @status:impl/done
  4. @fact:RATCHET-ORPHAN-BACKLOG orphan backlog — untagged exports the ratchet will block on: tag the export (@implements spec://…), @scope its file, or move it out of the public surface. @status:impl/done

Tier 2 — weekly {#tier-two}

  • @fact:WEEKLY-FAST-LOOP typescript-ai-native fast-loop — every cell answers inside the budget; a cell with NO tests fails the check (the loop must exist). @status:impl/done
  • @fact:WEEKLY-TRIPWIRE typescript-ai-native tripwire --base origin/main — debt that this week's changes touched; each fired entry is addressed in the PR text: pulled-in, re-dispositioned, or consciously deferred. @status:impl/done
  • @fact:WEEKLY-REREAD-DISABLE-AND-EXEMPT-LISTS Re-read the floor_disable list and the [[typescript.exempt]] list: does each reason still hold? @status:impl/done

Output contract {#output-contract}

@fact:END-EVERY-SWEEP-WITH-THE-OUTCOME-TABLE End every sweep with an outcome table — this stack's own addition, and not something the Sweep Playbook prescribes: per tier, what ran, what was found, the ONE ratchet item taken, and what was deliberately left (with why). The Playbook's own closing contract is its §4 «Output of a sweep» — topic-grouped commits, the refreshed health snapshot, the resume pointer — and it holds here unchanged. @status:impl/done

@fact:green-gates-only-is-not-a-sweep A sweep that only reports green gates did the floor's job, not the sweep's. @status:spec/done

The generation-time assistant (before you edit, not instead of the floor) {#generation-time-assistant}

@fact:STACK-SHIPS-AN-AGENTIC-TYPE-ORACLE The stack ships an agentic type oracle. @status:impl/done

@fact:check-the-hypothetical-content-lead Before writing a nontrivial .ts edit, check the HYPOTHETICAL content instead of paying a red floor iteration: @status:impl/done

sh
vibe bin exec typescript-ai-native-tcg -- validate src/cells/<cell>/index.ts \
    --content-from - --root .   # the edit on stdin; exit 1 = would fail

@fact:MCP-ALTERNATIVE or, when the vibevm MCP server is mounted, call tcg_validate with the content argument (plus tcg_scope / tcg_complete / tcg_type for in-scope symbols, type-valid completions, and quick info). @status:impl/done

@fact:RESPONSES-CARRY-THE-SAME-CONFORM-FINDINGS Responses carry the SAME conform findings as the gate, flagged baselined or new, with guide-citing advice — a new finding in the answer means the floor WILL go red if you write that edit. @status:impl/done

@fact:FLOOR-STAYS-THE-TRUTH The floor stays the truth; the oracle exists so you reach it green on the first try. @status:impl/done

Frequently asked questions

What does the Typescript Ai Native Sweep AI skill do?

Run the recurring AI-Native discipline sweep on this TypeScript project — the seven-step floor first, then the health collector's ratchet items. Use daily or several times a day on an active tree; any single item is a safe stop.

Why use Typescript Ai Native Sweep on TypingMind?

Because you install it once and use it with any model. Typescript Ai Native Sweep is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Typescript Ai Native Sweep in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/vibevm/vibevm/tree/main/vibevm/vibepacks/org.vibevm.ai-native/typescript-ai-native-lang/v1.0.0/vibevm/vibespecs/skills/typescript-ai-native-sweep. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Typescript Ai Native Sweep?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Typescript Ai Native Sweep?

As many as you like. As long as a model supports skills, you can use Typescript Ai Native Sweep with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Typescript Ai Native Sweep AI skill free?

It is published on GitHub by vibevm. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇