← Back to Skills Marketplace
441
Downloads
1
Stars
3
Active Installs
12
Versions
Install in OpenClaw
/install truth-tutor
Description
Diagnose why a learner does not understand a topic, paper, concept, or explanation, then deliver blunt feedback plus an actionable improvement plan. Includes...
Usage Guidance
This skill appears coherent and low-risk: it only uses its bundled guidance files and asks the user for context. Before installing, consider: (1) the skill's default tone is blunt/strict — if you're sensitive, instruct it to use a softer strictness level or decline to use 'brutal' mode; (2) do not paste private, proprietary, or personally identifying documents into the conversation (the skill asks for paper excerpts and prior answers); (3) because it is instruction-only, there is no hidden install activity or credential requests; (4) test the skill with a non-sensitive example first to verify its behavior and tone. If you need the skill to operate differently (softer voice, safety checks), ensure you include that preference with each request.
Capability Analysis
Type: OpenClaw Skill
Name: truth-tutor
Version: 1.0.9
The 'truth-tutor' skill is a purely instructional framework designed to guide an AI agent in providing diagnostic educational coaching. It contains no executable code, shell commands, or network activity. The skill focuses on identifying learning gaps (e.g., prerequisite or terminology gaps) and provides structured templates for feedback. It includes explicit safety boundaries in SKILL.md to prevent the agent from becoming abusive or encouraging self-harm, and it does not attempt to access sensitive system data or exfiltrate information.
Capability Assessment
Purpose & Capability
Name/description (diagnose learning gaps and prescribe fixes) match the runtime instructions and included reference files. All declared capabilities are about diagnosing/repairing learning; nothing requires cloud creds, binaries, or unrelated privileges.
Instruction Scope
SKILL.md instructs the agent to collect only user-provided context (topic, what they know, paper/title, prior answers). It references local helper files for taxonomy and templates. There are no instructions to read system files, environment variables, or to call external endpoints.
Install Mechanism
Instruction-only skill with no install spec and no code files; nothing will be written to disk or downloaded during install.
Credentials
No environment variables, credentials, or config paths are requested (appropriate). Note: the skill will ask users to paste content such as paper excerpts or prior chat answers — users should avoid sharing sensitive or private data.
Persistence & Privilege
always is false and the skill does not request persistent system privileges or modify other skills. It allows normal autonomous invocation (platform default), which is expected for a user-invocable coaching skill.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install truth-tutor - After installation, invoke the skill by name or use
/truth-tutor - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.9
P0+P1: Ontology integration, multi-model diagnosis, automation workflows
v1.0.7
Fix: web server port fallback when default port 3474 is already in use (avoid EADDRINUSE).
v1.0.6
Version 1.0.6
- No file or documentation changes detected in this release.
- Internal version update only; functionality and guidance remain unchanged.
v2.0.0-beta
No user-visible changes in this version.
- Skill version updated to 2.0.0-beta, but no file changes detected.
- All functionality and documentation remain unchanged from previous release.
v1.0.5
No file changes were detected in this version.
- No updates or modifications to code or documentation.
- Functionality and behavior remain unchanged from the previous version.
v1.0.4
No changes detected in this version.
- No file or documentation updates between 1.0.4 and the previous version.
- Behavior, features, and user guidance remain unchanged.
v1.0.3
No visible changes in this version.
- No file modifications or content updates detected.
- Version incremented to 1.0.3, but all documentation and logic remain identical to the previous release.
v1.0.1
- No visible changes in this version; content and functionality remain the same.
- Version number updated from 0.4.0 to 1.0.1.
v1.0.0
Initial public release.
- Provides diagnosis-first coaching to identify learning gaps before offering solutions.
- Supports general, paper-reading, and alphaXiv recovery modes for targeted feedback.
- Includes strictness levels to match user preference for feedback tone.
- Emphasizes dense, actionable reports and avoids generic praise or vague encouragement.
- Enforces clear safety guidelines to maintain ethical critique.
v0.2.0
Minimal paper-reading Web UI: left PDF, right conversation, top-only controls. Also includes local MiniMax fallback and web workflow improvements.
v0.1.1
Add dedicated paper-reading and alphaXiv modes, new docs/examples/tests, and improved CLI flows.
v0.1.0
Initial release: diagnosis-first learning coach for paper reading and skill building.
Metadata
Frequently Asked Questions
What is Truth Tutor?
Diagnose why a learner does not understand a topic, paper, concept, or explanation, then deliver blunt feedback plus an actionable improvement plan. Includes... It is an AI Agent Skill for Claude Code / OpenClaw, with 441 downloads so far.
How do I install Truth Tutor?
Run "/install truth-tutor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Truth Tutor free?
Yes, Truth Tutor is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Truth Tutor support?
Truth Tutor is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Truth Tutor?
It is built and maintained by Zhu Keqi (@sii-k7); the current version is v1.0.9.
More Skills