Guide

Every page in the engine’s docs/guide/ directory.

• Change Radar: what changed? — “It worked last week” is the most common sentence in agent operations and the hardest to act on. Somebody edited the system prompt. A tool was added. The provider rolled a new…
• Client libraries — Two thin clients for the routes an agent harness talks to — the run fuses, replays and forks, diffs, recall, and the ledger — live in this repository, with tape-test helpers on…
• Context X-Ray and Cache Autopsy — Traza already stores every request an agent sent to a model. This guide covers reading one of those requests as what it was — the context window the model saw — and asking it the…
• Data model — Traza stores one kind of record: a span. Everything else — traces, sessions, token rollups, waterfalls — is derived from spans at read time.
• Foresight: the Fate Map and predictive fuses — Traza holds every step of every run an agent made. Read one at a time, those runs are traces. Read together, they say something no single trace can: which paths an agent takes,…
• Fork & Rewind, and tape tests — A replay re-runs an agent against its own recording. This guide covers what you can do once a replay can be told where the recording stops being the truth:
• Getting started — From nothing to a span you can see in the browser. Nothing else needs to be installed: no database, no queue, no mandatory container.
• HTTP API reference — Every route Traza serves. Responses are JSON unless noted. All examples were run against a live server; the response bodies shown are real output.
• Ingest — Traza accepts spans on two routes. They write to the same store, honour the same (traceid, spanid) primary key, and return under the same durability contract. The only difference…
• Recording coding agents: Claude Code and Codex, zero setup — Claude Code and the Codex CLI already write down everything they do: one JSONL transcript per session, on your own disk. traza-server import turns those transcripts into runs —…
• Replay: tapes, the replay proxy, diffs and recall — Traza already records what an agent did. This guide covers what you can do with that recording afterwards:
• Run fuses — A fuse holds an agent run to limits — cost, tokens, steps, LLM calls, tool calls, duration, repeated steps, consecutive failures — and gives the run a verdict:
• Steer: live runs, nudges and topology — The run guard already follows every live agent run in memory — its usage, its verdict — and is asked before each step whether the step may run. Steering puts a person in that loop:
• The ledger and evidence packs — A generation’s manifest proves what the store holds now: every load-bearing file, with its digest. It cannot prove what the store held last quarter, or that today’s manifest…
• The MCP server — Traza serves the Model Context Protocol at POST /v1/mcp, so a coding agent can read this store the way the dashboard does — searching spans, opening traces, grouping failures,…
• The trace browser — Traza ships a dashboard: a React single-page app that talks to the same public /v1 API any client uses. Twenty-seven screens, grouped by the question you arrived with — what…
• TQL: query agent behaviour like a regex — Traza holds every step of every run an agent made. TQL (trajectory query language) asks whether a run’s steps contain a shape: a failing test, then an edit to a test file, then a…