{% extends "base.html" %} {% block head %} {% endblock %} {% block content %}
🤖 AI Model
Primary model i
Primary model
The AI model used for all responses. DuckClaw uses LiteLLM, so any supported provider works. Prefix with groq/, gemini/, or ollama/ for non-Claude models. Example: gemini/gemini-2.0-flash
Max tokens i
Max tokens
Maximum tokens the model generates per response. Higher = longer answers but more API cost. Range: 25632768. Recommended: 4096.
Temperature i
Temperature
Controls randomness. 0 = deterministic and precise. 1–2 = more creative and varied. Recommended: 0.7 for general chat.
Cost tracking i
Cost tracking
Estimates token cost for each LLM call and logs it. Useful for monitoring API spend. Works best with Claude and OpenAI models.
🛡️ Permissions
Default tier i
Default permission tier
Fallback tier for unclassified actions.
🟢 safe — silent auto-approve
🔵 notify — auto-approve + notification
🟡 ask — require your confirmation
🔴 block — never allow
Audit log i
Audit log
Records every action DuckClaw takes — skill calls, file reads, web searches — to a persistent log. Required for the Audit Log page to work.
Notify on safe i
Notify on safe
Show a notification even for SAFE-tier actions (which are normally silent). Turn on for full visibility into everything DuckClaw does.
🔒 Security
Prompt injection defense i
Prompt injection defense
Scans all incoming messages and LLM responses for injection patterns — attempts by malicious content to override DuckClaw's instructions. Strongly recommended to keep enabled.
Context isolation i
Context isolation
Wraps external data (web results, file contents) in a security fence so the LLM treats it as data only — not as instructions. Prevents indirect prompt injection attacks.
🖥️ Dashboard
Port i
Dashboard port
The local port where the DuckClaw dashboard is served. Default: 8741. A restart is required after changing this value.
🧠 Memory read-only
Database i
SQLite database path
Stores all user facts, conversation history, and ingested file records. Editable in the Database page.
{{ config.memory.db_path }}
Vector index i
ChromaDB vector index path
Stores semantic embeddings of conversations and ingested documents for memory search. Used when DuckClaw retrieves relevant past context.
{{ config.memory.chroma_path }}
Max facts i
Max facts
Maximum number of user facts held in memory at once. Older facts are pruned when the limit is reached to keep the context window lean.
{{ config.memory.max_facts }}
Restart DuckClaw to apply changes.
{% endblock %} {% block scripts %} {% endblock %}