← Back to Skills Marketplace
Lemma
by
deepak-jha-kgp
· GitHub ↗
· v1.0.0
· MIT-0
124
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install lemma
Description
Lemma is an AI operating system platform for business teams. Use this as the single entrypoint when designing, provisioning, testing, or improving Lemma pods...
README (SKILL.md)
Lemma
Use this as the main skill for the Lemma package.
This package is organized as one parent skill with domain modules under modules/.
Lemma helps teams build AI-powered operating systems around real business workflows. Each pod combines data, logic, orchestration, and operator experiences in one bounded system.
Required Routing Step
Before implementation, pick the module guide that matches the task and follow it.
- Platform design and sequencing:
modules/lemma-main/GUIDE.md - Integrations and operation discovery:
modules/lemma-integrations/GUIDE.md - Functions and backend patterns:
modules/lemma-functions/GUIDE.md - Datastore design and seed data:
modules/lemma-datastores/GUIDE.md - Workflow orchestration:
modules/lemma-workflows/GUIDE.md - Desk app architecture and implementation:
modules/lemma-desks/GUIDE.md - Assistant behavior and configuration:
modules/lemma-assistants/GUIDE.md - Agent design and task execution:
modules/lemma-agents/GUIDE.md - Workspace execution and troubleshooting:
modules/lemma-workspace/GUIDE.md - Inline widget work:
modules/lemma-widget/GUIDE.md
Cross-Module Convention
Use shared CLI drift notes from:
modules/lemma-main/references/known-cli-behavior.md
Build Order Guardrail
For non-trivial end-to-end delivery, follow:
- integrations
- functions
- workflows
- desks
Do not start desk action wiring until upstream verification is green.
Usage Guidance
This package looks like in-repo product documentation and developer tooling for a Lemma platform, not a small single-purpose helper — treat it as source code you will run. Before installing or executing anything: 1) Inspect the included shell scripts (especially init-artifact.sh and bundle-artifact.sh) to confirm they do only expected, local operations and do not call external URLs you don't trust. 2) Do not paste production LEMMA_TOKEN or other high-privilege credentials into workspace previews or browser localStorage; use a least-privilege, short-lived test token and rotate it afterward. 3) Confirm the required environment variables and their minimum needed scopes — the skill's metadata claims none, but the docs expect many. 4) Run any scripts in an isolated/dev workspace (not on a machine with production secrets) and prefer to dry-run commands or read their contents first. 5) If you need a stronger assurance, provide the full contents of the larger scripts for a deeper review; if those scripts make outbound network calls or upload artifacts, consider the skill suspicious until those calls are audited.
Capability Analysis
Type: OpenClaw Skill
Name: lemma
Version: 1.0.0
The skill bundle provides a comprehensive framework for an AI agent to design, provision, and manage 'Lemma' pods, which are described as AI-powered business operating systems. It includes shell scripts for scaffolding React/Vite applications (`init-artifact.sh`) and bundling them (`bundle-artifact.sh`), alongside extensive markdown documentation (`GUIDE.md` files) that serves as a structured instruction set for the agent. While the scripts handle sensitive environment variables (e.g., `LEMMA_TOKEN` in `print-browser-auth-setup.sh`) and the agent is guided to use powerful workspace capabilities like shell execution and browser automation, these actions are strictly aligned with the platform's stated purpose of software development and deployment. No evidence of malicious intent, data exfiltration, or harmful prompt injection was found.
Capability Assessment
Purpose & Capability
The skill's name and description match the included module guides and developer workflows, but the repository contains runtime scripts and many guides that assume environment variables (LEMMA_TOKEN, LEMMA_BASE_URL, LEMMA_POD_ID, VITE_LEMMA_* etc.) and browser token injection. The registry metadata claims no required env vars or credentials, which is inconsistent with the guidance in the docs. That mismatch is unexpected and should be justified.
Instruction Scope
SKILL.md routes the agent to many module GUIDE.md files that instruct running CLI commands, starting long-lived shells, curling LEMMA_BASE_URL, injecting auth tokens into browser localStorage, using playwright-cli, and running included shell scripts (e.g., ./modules/lemma-desks/scripts/init-artifact.sh). The instructions therefore reference reading and using workspace env vars and performing network calls — actions outside a strictly read-only documentation skill. The instructions do not declare limits on using tokens or where to send data, and they instruct browser-local token injection which can expose credentials if preview URLs are public.
Install Mechanism
There is no install spec (instruction-only), which is low risk for automatic installs. However, the package includes several shell scripts (notably init-artifact.sh and bundle-artifact.sh). Because the skill bundle contains executable scripts, an agent that follows the guides could run them; the package does not declare what those scripts do in the SKILL.md. Absence of an installation step reduces supply-chain risk, but embedded scripts still deserve inspection before execution.
Credentials
The skill declares no required environment variables or credentials, yet the module guides repeatedly assume the presence of sensitive env vars (LEMMA_TOKEN, LEMMA_BASE_URL, LEMMA_AUTH_URL, LEMMA_POD_ID and VITE_LEMMA_*). The guides also instruct injecting tokens into browser localStorage for testing. Requesting or assuming these secrets without declaring them is disproportionate and opaque — users should confirm which credentials are actually needed and avoid using high-privilege or production tokens for testing.
Persistence & Privilege
Flags show always:false and default autonomous invocation allowed. The skill does not request persistent inclusion or modify other skills. This is normal; no elevated platform privileges are requested in the metadata.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install lemma - After installation, invoke the skill by name or use
/lemma - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of the lemma skill—an AI operating system platform for business teams.
- Provides a single entrypoint for designing, provisioning, testing, and improving Lemma pods, integrations, functions, workflows, and more.
- Introduces a modular structure with specific guides for each domain (e.g., integrations, datastores, agents, desks).
- Establishes a required routing step: pick the relevant module guide before implementation.
- Documents conventions for cross-module work and references shared CLI behavior notes.
- Outlines a recommended build sequence for complex workflows to ensure upstream verification before progressing.
Metadata
Frequently Asked Questions
What is Lemma?
Lemma is an AI operating system platform for business teams. Use this as the single entrypoint when designing, provisioning, testing, or improving Lemma pods... It is an AI Agent Skill for Claude Code / OpenClaw, with 124 downloads so far.
How do I install Lemma?
Run "/install lemma" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Lemma free?
Yes, Lemma is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Lemma support?
Lemma is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Lemma?
It is built and maintained by deepak-jha-kgp (@deepak-jha-kgp); the current version is v1.0.0.
More Skills