← Back to Skills Marketplace
Adaptive Reasoning
by
MarjorieBroad
· GitHub ↗
· v1.0.0
· MIT-0
67
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install mar-adaptive-reasoning
Description
Automatically assess task complexity and adjust reasoning level. Triggers on every user message to evaluate whether extended thinking (reasoning mode) would...
Usage Guidance
This skill is coherent and low-risk technically (no downloads, no credentials). Before enabling it, consider: 1) It is designed to automatically turn on extended reasoning for messages it scores as complex — you may get longer, slower, and more token-expensive replies without an explicit prompt. 2) It instructs the agent to silently activate reasoning and append emoji (🧠 / 🧠🔥) to responses; if you prefer explicit user consent or no visible markers, disable or modify the behavior. 3) The SKILL.md references a session_status tool and '/reasoning' command — verify your agent runtime supports those commands or the skill's activation directives will be no-ops. If you want strict control, keep the skill user-invocable only (don’t set it to run automatically) or manually invoke reasoning when needed.
Capability Analysis
Type: OpenClaw Skill
Name: mar-adaptive-reasoning
Version: 1.0.0
The skill bundle is a meta-cognitive framework designed to help an AI agent self-assess task complexity and toggle extended reasoning modes. The instructions in SKILL.md are focused entirely on internal state management (e.g., using '/reasoning on' or the 'session_status' tool) and response formatting (appending emojis). There is no evidence of data exfiltration, malicious execution, or unauthorized access to sensitive system resources.
Capability Assessment
Purpose & Capability
Name/description (assess complexity and adjust reasoning) match the SKILL.md instructions. No unrelated binaries, env vars, or installs are required, which is proportionate to the stated goal.
Instruction Scope
Instructions remain within the claimed purpose (scoring heuristics, when to enable/disable reasoning, UI cues). However, the skill explicitly instructs the agent to 'Do not ask. Just activate.' and to use a session_status tool or '/reasoning on' command. That is a behavioral/policy directive (automatic enabling of extended reasoning and appending icons) rather than a request for external resources; it could affect UX or surprise users if they expect explicit consent.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest-risk install surface. Nothing is written to disk and nothing is downloaded.
Credentials
No environment variables, credentials, or config paths are requested. The skill does not ask for unrelated secrets or access.
Persistence & Privilege
The skill recommends triggering on every user message and toggling session-level reasoning state. It does not set always:true and does not request system-wide config changes, but it intends broad per-message behavior that will change how the agent answers (longer/thoughtful responses, emoji appended). Consider whether you want that global behavioral change.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install mar-adaptive-reasoning - After installation, invoke the skill by name or use
/mar-adaptive-reasoning - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of adaptive reasoning skill.
- Automatically assesses each user request for complexity and dynamically adjusts reasoning depth.
- Introduces a mental scoring system (0-10) based on features like multi-step logic, ambiguity, and novelty.
- Activates extended thinking and displays visual indicators (🧠, 🧠🔥) depending on complexity score.
- Includes clear thresholds and examples for when to escalate or lower reasoning level.
- Manual override commands and auto-downgrade logic included for efficiency.
Metadata
Frequently Asked Questions
What is Adaptive Reasoning?
Automatically assess task complexity and adjust reasoning level. Triggers on every user message to evaluate whether extended thinking (reasoning mode) would... It is an AI Agent Skill for Claude Code / OpenClaw, with 67 downloads so far.
How do I install Adaptive Reasoning?
Run "/install mar-adaptive-reasoning" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Adaptive Reasoning free?
Yes, Adaptive Reasoning is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Adaptive Reasoning support?
Adaptive Reasoning is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Adaptive Reasoning?
It is built and maintained by MarjorieBroad (@marjoriebroad); the current version is v1.0.0.
More Skills