Add Rules logo

Add Rules

OrganizationPopular
volcengine
add-rules

Compile durable constraints into OpenViking Rules with rules/compile. Use when the user asks to add rules or import a policy. Do not use for preferences, facts, or rule search.

Overview

Publishervolcengine
RepositoryOpenViking
Skill nameadd-rules
Stars
41.2K
Forks
3.3K
Bundled files
Instructions only
LicenseAGPL-3.0
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 volcengine on GitHub. Read the source before you install it.

Installation

Install the Add Rules 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/volcengine/OpenViking/tree/main/examples/skills/add-rules
  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/volcengine/OpenViking.git /tmp/OpenViking
mkdir -p .claude/skills
cp -r /tmp/OpenViking/examples/skills/add-rules .claude/skills/add-rules
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Add Rules 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 Add Rules 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 Add Rules 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.

Add Rules

Goal

Use POST /api/v1/rules/compile to compile user-authorized durable constraints into OpenViking Rules, then verify the task result and saved Rules.

Workflow

  1. Identify the user's rule text or policy sources and the intended OpenViking connection. Use the same instance and identity for upload, compilation, task polling, and verification.

  2. Choose exactly one invocation mode for the entire workflow:

    • Python SDK mode: use when the Agent runs in Python and the OpenViking SDK is available. Use SDK methods for source upload, compilation, task polling, and reading.
    • HTTP mode: use for other languages or general-purpose Agents. Use HTTP endpoints for source upload, compilation, task polling, and reading. Do not mix SDK calls and direct HTTP calls within one compilation workflow.
  3. Build a request with top-level instruction and sources; at least one must be non-empty. Preserve the source's obligations, applicability conditions, exceptions, thresholds, and identifiers without inventing policy.

  4. Continue only with the selected mode and use its exact parameter shape:

    Python SDK mode

    options supports sources. Each source may be a local file or directory path, a remote URL, a Resource URI, or an advanced source dictionary. The SDK uploads local paths automatically.

    python
    task = client.rule_compile(
        instruction="Compile the supplied policy into durable Rules.",
        options={
            "sources": [
                "/local/path/policy.md",
                "https://example.org/policy.md",
                "viking://resources/team-policy",
            ]
        },
    )
    task_id = task["task_id"]

    Poll with client.get_task(task_id) and read saved content with client.read_raw(uri).

    HTTP mode

    HTTP has no options wrapper. Put instruction and sources directly in the JSON body. Use resource_uri for an existing Resource, path for a remote text source, or temp_file_id returned by POST /api/v1/resources/temp_upload for local content.

    json
    {
      "instruction": "Compile the supplied policy into durable Rules.",
      "sources": [
        {"resource_uri": "viking://resources/team-policy"},
        {"path": "https://example.org/policy.md"},
        {"temp_file_id": "tmp_xxx", "source_name": "policy.md"}
      ]
    }

    Submit this body to POST /api/v1/rules/compile. Never send a client-local path to that endpoint. Poll GET /api/v1/tasks/{task_id} and read saved content through GET /api/v1/content/read?uri=...&raw=true.

    Save the returned task_id; acceptance does not mean compilation has completed.

  5. Poll using the selected mode until the task is completed, failed, or cancelled. Query the existing task rather than resubmitting the request.

  6. Inspect result.operations, written_uris, and write_result. Distinguish saved Rules from NOOP, UNCERTAIN, and REJECT results; completed does not guarantee new Rules.

  7. Read representative saved Rules and their Scenarios using the selected mode. Check the policy text and evidence, and verify rules/<scenario_id>/abstract.md holds Scenario metadata while Rule files sit directly beside it as <atomic_id>.md. Report the task ID, actual changes, representative URIs, and unresolved decisions.

Boundaries

  • Compile durable constraints only; do not promote preferences, facts, or one-task advice.
  • Prefer faithful, non-empty applies_when conditions supported by the source. Empty applies_when means applicable in every situation except exception, not unknown.
  • Request lifecycle changes only when the user explicitly authorizes them. Use ARCHIVE only when the user's instruction explicitly asks to delete, discard, or move a Rule to the recycle bin. Use PAUSE for pause, disable, deactivate, retire, stop-applying, or stop-using requests without deletion intent. Archived Rules are logically deleted: they are not compiler candidates and cannot be restored through Rule compilation.
  • Use client connection configuration, not service-side model credentials. Do not switch instances or expose credentials to resolve a failed request.
  • The API accepts at most 16 sources and 64000 instruction characters; source collection reads at most 128 files and 128000 characters. Split oversized material into self-contained policies with their conditions and exceptions rather than silently omitting content.
  • Rule compilation requires readable text. Do not submit image, audio, or video sources.
  • If a request fails or its status is unknown, check the task and saved files before retrying. Previously imported Resources are not rolled back when later Rule compilation fails.
  • Do not treat an immediate empty search result as proof that writing failed; semantic indexing may still be pending.

Frequently asked questions

What does the Add Rules AI skill do?

Compile durable constraints into OpenViking Rules with rules/compile. Use when the user asks to add rules or import a policy. Do not use for preferences, facts, or rule search.

Why use Add Rules on TypingMind?

Because you install it once and use it with any model. Add Rules 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 Add Rules in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/volcengine/OpenViking/tree/main/examples/skills/add-rules. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Add Rules?

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 Add Rules?

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

Is the Add Rules AI skill free?

Yes. It is published on GitHub by volcengine under the AGPL-3.0 license. 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 👇