Route agent execution to LLMChatEngine behaviors by exec_mode.
npx skills add https://github.com/joshka0/foxctl --skill foxctl-daemon
Please help me install this Agent Skill. Skill: foxctl Daemon Source: https://github.com/joshka0/foxctl/tree/main/configs/skills/foxctl-daemon Command: npx skills add https://github.com/joshka0/foxctl --skill foxctl-daemon
It helps you understand and control how foxctl’s agent daemon selects execution engines so reactive agents respond correctly and autonomous/proactive agents can run tool-calling loops reliably.
reactive, autonomous, and proactive modes using an internal companion service.Spawn an autonomous agent that self-runs with tool calling by executing: foxctl agent spawn --name "Coder" --role coder --exec-mode autonomous.
High-intent search queries and answers about installing and using this skill.
Agent execution mode routing dispatches reactive, autonomous, and proactive behaviors to the correct LLMChatEngine using deterministic exec_mode selection. This ensures reactive agents respond to events while autonomous agents run continuous tool-calling loops reliably.
Autonomous agents operate via mailbox polling to wait for work, while proactive agents self-initiate background or scheduled tasks. Both map to distinct operational modes that determine how the LLM engine manages tool-calling loops and multi-step reasoning.
Context budget limits are enforced using max-context-tokens configuration, which triggers stop reasons like StopReasonContextBudget. This halts agent execution when token limits are reached, ensuring sessions stay within defined operational boundaries.
Provider defaults like cerebras fallback can be configured for LLM engine selection in agent daemons. Model selection matches each agent's intent, ensuring the correct provider resolves based on the defined execution mode and operational requirements.
Spawn an autonomous agent with tool calling by executing foxctl agent spawn with the name, role, and exec-mode set to autonomous. This initializes the agent to self-run tool-calling loops using the routed LLMChatEngine behavior.
Session continuity persists agent turns and links sessions to maintain context across multi-step reasoning loops. This allows autonomous and proactive agents to resume execution reliably without losing prior tool-calling state or operational context.
zircoteCreate, configure, and shut down autonomous agent teams with permissions and messaging.
ijindal1Design and implement tool-using agents with OpenAI models.
eveldCoordinate specialized sub-agents across five implementation phases with token budgeting.
limit5Route HD phase AI companion chat queries to appropriate agents or workflows.