We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
Document reasoning stream support
Add Goals and Todos page; link from sidebar, Graph Runtime, Examples Document the per-thread goal (graph::goals) and task board (graph::todos) primitives in depth: data models, Store-backed persistence, the harness tool surfaces, and the graph-native continuation (goal_gate_node self-driving loop, run_continuation_tick driver, note_user_turn). Add a Graph Runtime section and an Examples entry for the goals_and_todos example, and link the new page from the sidebar.
Drop crate version numbers from the wiki and fix stale feature notes Remove pinned crate versions (1.0.0 / v1.2.x prose, `tinyagents = "1"` Cargo snippets → version-less `cargo add`) so the docs don't go stale on every release. Also correct two spec mismatches the feature removal left behind: Quick-Start still listed `openai` as an opt-in Cargo feature, and Providers claimed hosted providers were feature-gated — both now reflect the OpenAI adapter being compiled in by default.
Reflect OpenAI provider being compiled in by default Drop --features openai from example/test commands and remove openai-feature prose across Providers, Examples, Quick-Start, Development, and Expressive Language pages.
Add agent-friendly Capabilities index and document v1.2.x SDK-gap features New Capabilities.md master index (per-surface capability tables + Recently-added section + agent quick-start decision list) linked from Home and the sidebar. Harness, Graph-Runtime, and Registry pages document the new tool-policy enforcement, unknown-tool recovery, workspace isolation, tool-exposure auditing, middleware control precedence, budget reservation/reconcile, stable event ids, map/reduce timeouts/cancellation, orchestrate_list filters, registry introspection, and storage conformance. Examples gains runnable snippets per feature.
Enhance documentation across multiple files to reflect updates in the TinyAgents architecture. Key changes include: updating Architecture.md to include parallel-agent orchestration and observability in the graph runtime; revising Development.md to reflect the new versioning and optional backend features; expanding Examples.md with a new `subconscious_loop` example; and improving clarity in Home.md and Providers.md regarding setup and model capabilities. Additionally, updates to Graph-Runtime.md and Harness.md provide more detail on graph execution and observability features.
Enhance documentation for TinyAgents architecture, examples, and providers. Update Architecture.md to clarify the recursive language-model harness concept and its five surfaces. Expand Examples.md with detailed descriptions of runnable examples and their recursive concepts. Revise Graph-Runtime.md to elaborate on the graph execution model and node behavior. Improve Home.md and Providers.md for better clarity on setup and provider integration.
Document TinyAgents wiki