Usage: agents-exe [--api-keys AGENTS-KEY] [--log-file LOGFILE] [--log-http LOGHTTP] [--log-json-file JSONFILE] [--agent-file AGENTFILE] [-a|--agent SLUG] COMMAND [--set NAME=VALUE] [--pin NAME=VALUE] [--set-json NAME=JSON] [--pin-json NAME=JSON] [--params-file FILE] Your favorite agent framework. Available options: --api-keys AGENTS-KEY path to json-file containing API keys (default: "/home/lucas/.config/agents-exe/secret-keys") --log-file LOGFILE raw log file (default: "agents-logfile") --log-http LOGHTTP http log sink --log-json-file JSONFILE local JSON file log sink --agent-file AGENTFILE root agent(s) description files defaults to 'agent.json' multi-support vary from command to command -a,--agent SLUG Select agent by slug instead of file path --set NAME=VALUE Set a parameter value; repeatable --pin NAME=VALUE Set a parameter value that cannot be overridden; repeatable --set-json NAME=JSON Set a parameter to any JSON value; repeatable --pin-json NAME=JSON Set a JSON parameter value that cannot be overridden; repeatable --params-file FILE JSON file of {"name": value, ...} parameter values; repeatable. Flags win over a file's values. -h,--help Show this help text Available commands: check Validate agent configurations and optionally dump tool schemas check-tool-call Validate a tool call payload against a tool schema (reads JSON from stdin) config Configure agents-exe (git-config style) list-tool-calls List all tool calls from a session file replay-tool-call Replay a tool call from a session file, validating and optionally executing tui Interactive terminal UI, over an embedded runner or, with --attach, a running server spectate Read-only live dashboard of a running server: session tree, tool calls and model text run echo-prompt describe describe-tool Validate and describe a tool script without agent context init mcp-server session-print Print a session file in markdown format session-edit Edit a session file (reads JSON from STDIN, writes JSON to STDOUT) session-index Manage the session search index session-search Search session files with fuzzy matching paths Show important configuration paths cowsay Display a message with the agents-exe mascot spec Display specification documentation (e.g., bash-tools) new Create new agent or tool scaffolding session Operate durable sessions: start, step, pause, resume, pending, complete, run-isolated tool-call Call a tool from the first loaded agent with JSON payload from stdin serve Run agents over HTTP (like agents-server), loading agents-exe.cfg.json like the TUI does