The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS.
Rust agent-orchestrationai-agentsai-assistantdesktopllmlocal-firstmcppersonal-aiprivacyrustsecond-braintauri
docs(architecture): update source links to fork branch
Steven Enamakel79d808e1abaa
diff --git a/docs/gitbooks/en/developing/architecture/agent-harness.md b/docs/gitbooks/en/developing/architecture/agent-harness.mdindex 7787a587dd..ac5b7ea934 100644--- a/docs/gitbooks/en/developing/architecture/agent-harness.md+++ b/docs/gitbooks/en/developing/architecture/agent-harness.md@@ -93,7 +93,7 @@ Some settings remain runtime-wide for every agent (the live autonomy policy'sconfig sub-agents re-read); the crate README lists them under "Stillruntime-wide".-Every agent turn runs through the published [`tinyagents`](https://crates.io/crates/tinyagents) harness. There are three entry points: `Agent::turn`, the channel and CLI bus path, and `run_subagent`. All of them go through the adapter seam in [`crates/openhuman-core/src/agent/tinyagents/`](https://github.com/tinyhumansai/openhuman/tree/main/crates/openhuman-core/src/agent/tinyagents) (`TurnSpec::run`). Context trimming is `MessageTrimMiddleware`, and cancellation is the tinyagents steering channel. Policy stop hooks (budget, thread-goal and iteration caps) fire through a `StopHookMiddleware` ([`tinyagents/stop_hooks.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/agent/tinyagents/stop_hooks.rs)) that pauses the run on the first stop vote. The channel route forwards live `AgentProgress` like the chat route. The shared `TurnProgress` seam lives in `agent/session_host/tool_progress.rs`.+Every agent turn runs through the published [`tinyagents`](https://crates.io/crates/tinyagents) harness. There are three entry points: `Agent::turn`, the channel and CLI bus path, and `run_subagent`. All of them go through the adapter seam in [`crates/openhuman-core/src/agent/tinyagents/`](https://github.com/tinyhumansai/openhuman/tree/main/crates/openhuman-core/src/agent/tinyagents) (`TurnSpec::run`). Context trimming is `MessageTrimMiddleware`, and cancellation is the tinyagents steering channel. Policy stop hooks (budget, thread-goal and iteration caps) fire through a `StopHookMiddleware` ([`tinyagents/stop_hooks.rs`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/agent/tinyagents/stop_hooks.rs)) that pauses the run on the first stop vote. The channel route forwards live `AgentProgress` like the chat route. The shared `TurnProgress` seam lives in `agent/session_host/tool_progress.rs`.Multi-agent orchestration runs on the tinyagents graph layer, through `tinyagents-orchestration`'s `parallel` module, the host `spawn_parallel` adapter and the shared `graph::orchestration` `TaskStore` lifecycle primitives re-exported from [`orchestration/mod.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/agent/orchestration/mod.rs):@@ -486,7 +486,7 @@ Why the rules.Enforcement. Two layers:-1. Loader-time (static). [`agents::loader::validate_tier_hierarchy`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/agent/registry/agents/loader.rs) runs over the merged registry (built-ins + workspace TOMLs) and refuses to boot a registry that lists a same-tier or worker-with-subagents entry. Built-in archetypes are checked at compile-test time; user-shipped TOMLs are checked at workspace load.+1. Loader-time (static). [`agents::loader::validate_tier_hierarchy`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/agent/registry/agents/loader.rs) runs over the merged registry (built-ins + workspace TOMLs) and refuses to boot a registry that lists a same-tier or worker-with-subagents entry. Built-in archetypes are checked at compile-test time; user-shipped TOMLs are checked at workspace load.2. Runtime depth gate (dynamic). Independent of tier, the sub-agent runner caps total spawn chain depth at `MAX_SPAWN_DEPTH = 3` via a task-local counter incremented across `run_subagent`, surfaced as a `SpawnDepthExceeded` agent error. This makes a user-shipped TOML that drops the tier annotation still unable to recurse past three hops.> Status: the loader-time tier check, `agent_tier` field, and runtime depth-counter task-local are live. Depth is bounded by both the static loader contract and the runtime `MAX_SPAWN_DEPTH = 3` guard.diff --git a/docs/gitbooks/en/developing/architecture/flows-on-tinyagents.md b/docs/gitbooks/en/developing/architecture/flows-on-tinyagents.mdindex f42d0bcf4a..8ef46a4887 100644--- a/docs/gitbooks/en/developing/architecture/flows-on-tinyagents.md+++ b/docs/gitbooks/en/developing/architecture/flows-on-tinyagents.md@@ -67,13 +67,13 @@ flowchart LR4. outcome: a `RunOutcome { output, pending_approvals, cancelled }`; the hostpersists live steps through the `FlowRunObserver` and exports the durablegraph observations to Langfuse- ([`tinyflows/langfuse_export.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/tinyflows/langfuse_export.rs)).+ ([`tinyflows/langfuse_export.rs`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/flows/tinyflows/langfuse_export.rs)).Because the engine keys persisted state by a caller-supplied `thread_id`,durable human-in-the-loop resume is `engine::resume_with_checkpointer` over the same`tinyagents::graph::SqliteCheckpointer` the agent harness uses: opened once perhost at `<workspace_dir>/flows/checkpoints.db`-([`caps/ops.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/tinyflows/caps/ops.rs), `open_flow_checkpointer`).+([`caps/ops.rs`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/flows/tinyflows/caps/ops.rs), `open_flow_checkpointer`).## Run state: one JSON map, a merge reducer, and the `{json,text,raw}` envelope@@ -144,7 +144,7 @@ code, persistence and sub-workflow lookup are each a capability trait that the himplements ([`vendor/tinyflows/crates/tinyflows/src/caps/mod.rs`](https://github.com/tinyhumansai/tinyflows/blob/main/crates/tinyflows/src/caps/mod.rs)).`openhuman::flows::tinyflows::caps` supplies one adapter per trait, assembled into a`Capabilities` bundle per run by `build_capabilities`-([`caps/`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/tinyflows/caps/mod.rs)):+([`caps/`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/flows/tinyflows/caps/mod.rs)):| tinyflows trait | Node(s) it backs | OpenHuman adapter | Wraps || ------------------ | ------------------------------- | ---------------------------------- | ------------------------------------------------ |@@ -231,7 +231,7 @@ around the whole engine future`[autonomy]` tier through `SecurityPolicy::gate_decision` for that node's`CommandClass` (`http_request` → Network, `code` → Write, native `oh:` tools →their classified class) in `enforce_node_tier_gate`-([`caps/`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/tinyflows/caps/mod.rs)):+([`caps/`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/flows/tinyflows/caps/mod.rs)):- a `readonly` run is blocked at the network/code boundary and never dispatches;- a `supervised` run's `Prompt` decision is escalated by `gate_call_for_tier`@@ -301,7 +301,7 @@ pattern the flow's own `agent` nodes use:| Executor | (the engine) | `flows_run` / `flows_resume` | the capability seam above |Both agents live under-[`crates/openhuman-core/src/flows/agents/`](https://github.com/tinyhumansai/openhuman/tree/main/crates/openhuman-core/src/flows/agents) as+[`crates/openhuman-core/src/flows/agents/`](https://github.com/senamakel/openhuman/tree/epic/7964-channel-v2/crates/openhuman-core/src/flows/agents) asfirst-class registry agents (`agent.toml` + `prompt.md`), on the reasoning tierwith narrow, safety-reviewed tool belts. The builder is tool-assistedend-to-end: it proposes graphs (validate-only), dry-runs them against mock@@ -335,12 +335,12 @@ sequenceDiagram| ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------- || [`vendor/tinyflows/crates/tinyflows/src/`](https://github.com/tinyhumansai/tinyflows/tree/main/crates/tinyflows/src) | The engine: `model/`, `validate.rs`, `compiler.rs`, `engine.rs`, `expr.rs`, `caps/`, `nodes/`. || [`vendor/tinyflows/crates/tinyflows/src/caps/mod.rs`](https://github.com/tinyhumansai/tinyflows/blob/main/crates/tinyflows/src/caps/mod.rs) | The capability traits + `Capabilities` bundle. |-| [`crates/openhuman-core/src/flows/tinyflows/caps/`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/tinyflows/caps/mod.rs) | The host adapters, `build_capabilities`, `open_flow_checkpointer`, the two-layer gate helpers. |+| [`crates/openhuman-core/src/flows/tinyflows/caps/`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/flows/tinyflows/caps/mod.rs) | The host adapters, `build_capabilities`, `open_flow_checkpointer`, the two-layer gate helpers. || [`crates/openhuman-core/src/flows/ops.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/ops.rs) | `flows_run` / `flows_resume` / `flows_build` / `flows_discover` and CRUD. || [`crates/openhuman-core/src/flows/bus.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/bus.rs) | `FlowTriggerSubscriber`: the host-side multi-trigger bridge. || [`crates/openhuman-core/src/flows/builder_tools.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/builder_tools.rs) | The Composio-bound builder reads; the authoring tools are `tinyflows_copilot::tools`, bound by `flows/builder_host.rs`. |-| [`crates/openhuman-core/src/flows/agents/`](https://github.com/tinyhumansai/openhuman/tree/main/crates/openhuman-core/src/flows/agents) | `workflow_builder` + `flow_discovery` agent definitions. |-| [`crates/openhuman-core/src/flows/tinyflows/langfuse_export.rs`](https://github.com/tinyhumansai/openhuman/blob/main/crates/openhuman-core/src/flows/tinyflows/langfuse_export.rs) | Post-run export of the durable graph observations. |+| [`crates/openhuman-core/src/flows/agents/`](https://github.com/senamakel/openhuman/tree/epic/7964-channel-v2/crates/openhuman-core/src/flows/agents) | `workflow_builder` + `flow_discovery` agent definitions. |+| [`crates/openhuman-core/src/flows/tinyflows/langfuse_export.rs`](https://github.com/senamakel/openhuman/blob/epic/7964-channel-v2/crates/openhuman-core/src/flows/tinyflows/langfuse_export.rs) | Post-run export of the durable graph observations. |## See also