Changelog
New features, SDK improvements, and platform updates.
A clearer, more scannable layout: status, cost, and tags up front; a stat strip showing duration, event count, LLM calls, cost, and errors at a glance; a dedicated Graph tab; and a searchable, filterable timeline you can navigate with the keyboard. Event detail now leads with the answer, not the raw input.
Drag to pan and scroll to zoom instead of scrollbars. Every session graph now starts from a single visible node instead of scattered top-level steps, and tool calls show their real name — not a generic "tool" label.
log_tool(), log_decision(), and log_error() now accept a parent_event_id — nest an event under the LLM call that caused it, so the timeline and graph show real causal structure instead of a flat list of unconnected steps.
log_tool(..., parent_event_id=get_last_llm_event_id())log_tool() now accepts the same metadata your LLM calls already use — attach a latency and see it in the timeline and graph exactly like an LLM call's duration.
log_tool(..., metadata={"latency": seconds})Sessions now show which model they used — including when a session mixes models — derived automatically from your LLM calls. Fixed Gemini calls under-reporting cost despite correct token counts, added pricing for newer models, and models with no pricing data are now flagged instead of silently showing $0.
Live session updates no longer cut off early just because a session's status changed. Hardened backend request handling so brief overload is handled gracefully instead of cascading. A failed request to load your sessions list now shows a clear retry prompt instead of looking like an empty account.
Closed an edge case where a batch of events could be silently dropped during a brief network hiccup in the ingestion pipeline. Events are now retried until they're confirmed durable, instead of being lost if the hiccup landed at the wrong moment.
A new Metrics tab on Analytics discovers every set_metrics() key automatically and infers its type — score, rating, sentiment, boolean, or numeric. See distribution and trend per key, then group results by agent or any tag (model, connection, environment) to compare performance across your fleet.
session_ctx.set_metrics({...})start_session() no longer waits on network round-trips to fetch policies and pricing — both happen in the background now, and session end is fire-and-forget too. Startup latency is no longer coupled to Niitaka's response time, even on a slow connection.
start_session() — returns instantlySession state is now fully isolated per thread. Fixed a race condition where a background thread — an async LLM-judge call, a parallel tool run — could cross-contaminate session state with the main request thread under concurrent load.
LLM call inputs are now scrubbed of api_key, api_base, and other credential fields before they reach Niitaka. Local models (Ollama and friends) get accurate $0 cost tracking instead of being billed at fallback rates — and any unrecognised model now shows $0 rather than a guessed price.
Hardened backend connection handling against transient network drops, and fixed a rare error when viewing decision traces for background-thread sessions.
Tag any session with key-value pairs — connection name, team, region, user ID, or any dimension that matters to you. Tags appear as clickable chips in the Sessions table: click a chip to filter instantly. Stack multiple tag filters; the stat tiles update in real time to reflect only the tagged subset.
start_session(tags={...})No more defensive boilerplate. When api_key is not set, start_session() silently returns a no-op context manager and all logging calls are silent. Your agent code is identical whether Niitaka is configured or not — adopt it at your own pace or disable it in test environments.
start_session() returns no-op when unconfigured · niitaka.create_session()Tag each LLM call site in your agent with a named slot — "planner", "sql_generator", "interpreter". When a runtime config or experiment variant defines a prompt_slots override for that name, the correct system prompt is injected automatically — no code changes, no redeploy.
niitaka.prompt_slot("name")litellm.acompletion() is now captured alongside synchronous litellm.completion(). Async agents — FastAPI backends, asyncio pipelines, async LangChain chains — are fully observed with zero extra setup after calling instrument_litellm().
niitaka.instrument_litellm()Group related agents under slash-separated folder paths. Folders are created automatically on first use — no dashboard setup required. Navigate the hierarchy in the Agents view, or scope policies and runtime configs to an entire folder at once.
start_session(folder_path="team/subfolder")Full-text search across session goals, event content, tool inputs/outputs, decisions, and error messages — all from one search bar. PostgreSQL trigram indexes power fast partial-match results with highlighted snippets so you can spot the relevant line without opening each event.
Connect your Slack workspace directly from the Alerts settings page — no webhook URL hunting or manual channel configuration. OAuth scopes are minimal: Niitaka only requests permission to post to the channels you select. Existing webhook-based channels continue to work unchanged.