← Back to Skills Marketplace
Dynamic Temperature
by
Netanel Abergel
· GitHub ↗
· v1.0.2
· MIT-0
139
Downloads
0
Stars
1
Active Installs
3
Versions
Install in OpenClaw
/install dynamic-temperature
Description
Dynamic LLM temperature selection based on task type. Use when deciding what temperature to apply for a given task — scheduling, communication, creative writ...
README (SKILL.md)
Dynamic Temperature Skill
Purpose
Select the right LLM temperature for each task to balance precision and creativity. Lower = more deterministic. Higher = more creative/natural.
Temperature Scale
| Task Type | Temperature | Examples |
|---|---|---|
| Irreversible actions | 0.0 | Delete calendar event, send official email, destructive CLI ops |
| Scheduling / Commands | 0.2 | Meeting coordination, dates, facts, CLI commands |
| Analysis / Summaries | 0.3 | Status reports, structured thinking, meeting notes |
| General communication | 0.5 | Daily WhatsApp replies, updates, follow-ups |
| Briefings / Drafts | 0.6 | Morning briefing, drafting emails with warmth |
| Creative writing | 0.8 | Jokes, stories, icebreakers, tone-heavy content |
Decision Rule
Before generating any output, classify the task:
1. Is this an irreversible action (delete, send, post)?
→ temperature: 0.0
2. Is this scheduling, dates, or commands?
→ temperature: 0.2
3. Is this a summary or structured analysis?
→ temperature: 0.3
4. Is this a standard reply or update?
→ temperature: 0.5
5. Is this a briefing or warm message?
→ temperature: 0.6
6. Is this creative, funny, or expressive?
→ temperature: 0.8
When in doubt → 0.5
Per-Skill Recommendations
| Skill | Recommended Temp | Reason |
|---|---|---|
owner-briefing |
0.6 | Warm, readable, but structured |
meeting-scheduler |
0.2 | Precision required |
ai-meeting-notes |
0.3 | Factual summaries |
supervisor |
0.2 | Status facts only |
billing-monitor |
0.1 | Alerts must be accurate |
git-backup |
0.0 | No creativity needed |
self-learning |
0.4 | Reflective but grounded |
pa-eval |
0.3 | Analytical |
Implementation Notes
OpenClaw does not yet support per-message dynamic temperature natively. Until it does, apply this guide by:
- Setting temperature in your
agents.defaults.modelsconfig per model - Or noting the recommended temperature in each skill's
SKILL.mdfrontmatter - When spawning subagents for specific tasks, pass the appropriate temperature
Communication Override Rules (Temperature 0.0 absolute)
- Sending messages to people → always confirm before sending (irreversible)
- Deleting data → always confirm
- "sure thing" reply → exact string, no creativity, temperature 0.0
- Reaction signals (👍, ✅) → deterministic, no variation
Learned From
Training session between Heleni (Netanel's PA) and Selena (Daniel's PA), April 2026. Key insight from Selena: irreversible actions = 0.0, no exceptions.
Usage Guidance
This skill is an advisory guideline for choosing temperature and appears safe: it asks for nothing sensitive and installs nothing. Before relying on it, consider: (1) the skill is instruction-only and will only take effect when invoked—if you want these rules applied to every model call, put them in your agent core identity (SOUL.md) or implement them in platform configuration; (2) the guide recommends confirming irreversible actions at temperature 0.0 — ensure your agent actually enforces a confirmation step rather than auto-sending; and (3) verify the per-skill names in the table match your deployed skills so recommendations are applied where intended.
Capability Assessment
Purpose & Capability
The name/description match the content: the SKILL.md contains temperature rules and per-skill recommendations. Nothing in the files requires access to unrelated services or secrets.
Instruction Scope
The instructions stay on-topic (classify task type and choose a temperature). Implementation notes suggest updating agents.defaults.models or passing temperature to subagents — this touches agent configuration and runtime behavior but is reasonable for the stated goal. The guide also mandates confirmations for irreversible actions; users should verify that confirmation steps are actually implemented by the agent platform.
Install Mechanism
No install spec and no code files—this is instruction-only, so nothing is downloaded or written to disk by the skill itself.
Credentials
No required environment variables, credentials, or config paths are requested. The skill only recommends configuration changes and temperature values; it does not ask for secrets or external endpoints.
Persistence & Privilege
always is false and model invocation is allowed (default). The skill does not request permanent presence or modification of other skills' configurations. Note: DEPRECATED.md states this logic may be more appropriate in a core identity file (SOUL.md) if the user wants the rule applied universally.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install dynamic-temperature - After installation, invoke the skill by name or use
/dynamic-temperature - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
- Added "Communication Override Rules" section clarifying that certain communication tasks (e.g., sending messages, deleting data, exact replies) must use temperature 0.0 absolutely.
- Specified that reaction signals and set-phrase replies are always deterministic (temperature 0.0).
- No changes to recommendations or decision rules; documentation expanded for stricter application in sensitive communication scenarios.
v1.0.1
reactions rule, close-the-loop, reply-to rules; skill-master analytics hook; skill-analytics added
v1.0.0
- Initial release enabling dynamic selection of LLM temperature based on task type.
- Provides a clear temperature scale and decision rules for task classification (irreversible actions, scheduling, summaries, communication, and creativity).
- Includes per-skill temperature recommendations for commonly used assistant skills.
- Offers interim implementation guidance for platforms without native dynamic temperature support.
- Focuses on balancing precision and creativity to improve output quality and reliability.
Metadata
Frequently Asked Questions
What is Dynamic Temperature?
Dynamic LLM temperature selection based on task type. Use when deciding what temperature to apply for a given task — scheduling, communication, creative writ... It is an AI Agent Skill for Claude Code / OpenClaw, with 139 downloads so far.
How do I install Dynamic Temperature?
Run "/install dynamic-temperature" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Dynamic Temperature free?
Yes, Dynamic Temperature is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Dynamic Temperature support?
Dynamic Temperature is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Dynamic Temperature?
It is built and maintained by Netanel Abergel (@netanel-abergel); the current version is v1.0.2.
More Skills