mirror of
https://github.com/tinyhumansai/openhuman.git
synced 2026-07-27 21:08:00 +00:00
* feat: add initial project structure and documentation - Introduced the GNU General Public License (GPL) v3 in LICENSE file. - Added MCP configuration in .claude/mcp.json for server integration. - Created architecture documentation in docs/ARCHITECTURE.md outlining the platform's design and components. - Defined MVP specifications in docs/MVP.md for the Telegram-based Agent Assistant. - Established API reference for team management in docs/teams-api-reference.md. - Set up basic HTML structure in public/index.html and added logo image in public/logo.png. * feat: add initial project documentation and HTML structure - Introduced CODE_OF_CONDUCT.md to establish community guidelines and standards for behavior. - Created CONTRIBUTING.md to outline contribution process, development setup, and project conventions. - Added SECURITY.md to define the security policy, supported versions, and reporting procedures for vulnerabilities. - Established basic HTML structure in index.html for the application interface. * chore: remove hello-python skill files - Deleted skill.json and skill.py files for the Hello Python example runtime skill, as they are no longer needed in the project. * feat: port tinyhuman agent runtime from ZeroClaw into Tauri backend Port daemon supervisor, health registry, security (policy, secrets, audit, pairing), agent traits, and config modules from ZeroClaw (MIT) into a new tinyhuman/ module under src-tauri/src/. The daemon auto-starts on desktop and shuts down gracefully on app exit via CancellationToken. - health: global HealthRegistry with component tracking and JSON snapshots - security/policy: SecurityPolicy with command validation, risk levels, rate limiting - security/secrets: ChaCha20-Poly1305 SecretStore with legacy XOR migration - security/audit: AuditLogger with JSON-line events and log rotation - security/pairing: PairingGuard with brute-force protection and SHA-256 hashing - security/traits: Sandbox trait + NoopSandbox - config: minimal DaemonConfig with autonomy, reliability, secrets, audit sub-configs - daemon: supervisor with health state writer emitting Tauri events - agent/traits: Provider, Tool, Memory, Observer, RuntimeAdapter traits + Noop impls - commands/tinyhuman: Tauri commands for health, security policy, encrypt/decrypt - 185 inline unit tests across all modules - README updated with custom inference/tunneling/memory positioning Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> * feat: update README to reflect AlphaHuman Mk1 branding and enhanced description - Changed project title to "AlphaHuman Mk1" for clarity. - Revised project description to emphasize user-friendly AI capabilities and the use of the Neocortex Mk1 model. - Removed outdated sections on custom inference, tunneling, and memory, streamlining the content for better readability. * update readme * Port zeroclaw runtime into tinyhuman * Replace CLI mentions with UI language * Split gateway module into smaller units * Split channels and config schema modules * Fix tinyhuman build, tests, and tunnel integration * feat(tinyhuman): add missing modules and ui-friendly services * refactor: rename tinyhuman to alphahuman * chore: remove bottom text from Welcome component * feat(settings): add tauri command console * feat(daemon): enhance daemon mode handling and integrate rustls with ring feature * feat(settings): implement comprehensive configuration management in TauriCommandsPanel * refactor(TauriCommandsPanel): streamline error handling and enhance async function usage * feat(settings): add skill management functionality to TauriCommandsPanel * style(TauriCommandsPanel): update input styles for improved readability and user experience * feat(settings): add Skills and Agent Chat panels with navigation and integration management * feat(settings): implement browser access management in SkillsPanel and enhance AgentChatPanel with local storage functionality --------- Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
70 lines
2.0 KiB
Rust
70 lines
2.0 KiB
Rust
//! Agent and delegate agent configuration.
|
|
|
|
use schemars::JsonSchema;
|
|
use serde::{Deserialize, Serialize};
|
|
use std::collections::HashMap;
|
|
|
|
/// Configuration for a delegate sub-agent used by the `delegate` tool.
|
|
#[derive(Debug, Clone, Serialize, Deserialize, JsonSchema)]
|
|
pub struct DelegateAgentConfig {
|
|
/// Provider name (e.g. "ollama", "openrouter", "anthropic")
|
|
pub provider: String,
|
|
/// Model name
|
|
pub model: String,
|
|
/// Optional system prompt for the sub-agent
|
|
#[serde(default)]
|
|
pub system_prompt: Option<String>,
|
|
/// Optional API key override
|
|
#[serde(default)]
|
|
pub api_key: Option<String>,
|
|
/// Temperature override
|
|
#[serde(default)]
|
|
pub temperature: Option<f64>,
|
|
/// Max recursion depth for nested delegation
|
|
#[serde(default = "default_max_depth")]
|
|
pub max_depth: u32,
|
|
}
|
|
|
|
fn default_max_depth() -> u32 {
|
|
3
|
|
}
|
|
|
|
#[derive(Debug, Clone, Serialize, Deserialize, JsonSchema)]
|
|
pub struct AgentConfig {
|
|
/// When true: bootstrap_max_chars=6000, rag_chunk_limit=2. Use for 13B or smaller models.
|
|
#[serde(default)]
|
|
pub compact_context: bool,
|
|
#[serde(default = "default_agent_max_tool_iterations")]
|
|
pub max_tool_iterations: usize,
|
|
#[serde(default = "default_agent_max_history_messages")]
|
|
pub max_history_messages: usize,
|
|
#[serde(default)]
|
|
pub parallel_tools: bool,
|
|
#[serde(default = "default_agent_tool_dispatcher")]
|
|
pub tool_dispatcher: String,
|
|
}
|
|
|
|
fn default_agent_max_tool_iterations() -> usize {
|
|
10
|
|
}
|
|
|
|
fn default_agent_max_history_messages() -> usize {
|
|
50
|
|
}
|
|
|
|
fn default_agent_tool_dispatcher() -> String {
|
|
"auto".into()
|
|
}
|
|
|
|
impl Default for AgentConfig {
|
|
fn default() -> Self {
|
|
Self {
|
|
compact_context: false,
|
|
max_tool_iterations: default_agent_max_tool_iterations(),
|
|
max_history_messages: default_agent_max_history_messages(),
|
|
parallel_tools: false,
|
|
tool_dispatcher: default_agent_tool_dispatcher(),
|
|
}
|
|
}
|
|
}
|