turnstone-eval was misnamed: it was a prompt optimizer, not a measurement
harness. Split the 3252-line turnstone/eval.py into a strictly one-way
dependency (optimizer -> eval-core; core never imports the optimizer):
- turnstone/eval/core.py measurement substrate — everything up to and
including _run_iteration: provider detection, NullUI, HeadlessSession,
the test runner, score_run, aggregation, and neutral reporting.
- turnstone/eval/cli.py new measure-only `turnstone-eval` — the old
--no-optimize path promoted to the whole job (one _run_iteration call,
then print the summary table).
- turnstone/optimizer.py the UCB self-modify loop and its multi-agent
pipeline (analyst/optimizer/observer/diversifier/tool optimizer), now
`turnstone-optimizer`; imports from eval.core only.
- turnstone/eval/__init__.py re-exports the core public API for
back-compat (score_run, _match_action, _run_iteration, HeadlessSession).
_apply_tool_overrides lives in core (HeadlessSession needs it) rather than
alongside the other tree helpers, so the dependency stays one-way.
Breaking change: `turnstone-eval` now measures; use `turnstone-optimizer`
to optimize. Both code paths are behaviour-preserving — the moved function
bodies are byte-identical.
`man` and `math` duplicated capabilities already reachable through
`bash`; `plan_agent` is better expressed as a `task_agent` running a
planning skill, and carried a large amount of special-case machinery
(plan-review gate, refinement loop, per-kind model routing). Removing
all three shrinks the tool surface and cuts per-call token cost.
Also removed, as dead-once-the-tools-are-gone:
- the `math` sandbox executor (`turnstone.core.sandbox`) and its
`[sandbox]` extra; the eval analyst now runs bash-only
- the read-only `AGENT_TOOLS` sub-agent tool set and the `agent`
tool-metadata key (`task_agent`/`TASK_AGENT_TOOLS` retained)
- the plan-review protocol end to end: the `on_plan_review` UI hook,
`resolve_plan`, `POST /v1/api/plan` + `POST /v1/api/route/plan`,
the `plan_review`/`plan_resolved` SSE events, and their Python SDK /
TypeScript SDK / OpenAPI / frontend / Discord+Slack bindings
- the `model.plan_alias` / `model.plan_effort` settings and the
registry `plan_model` / `plan_effort` routing fields
TOOLS 31->28, TASK_AGENT_TOOLS 13->11; COORDINATOR_TOOLS unchanged.
BREAKING CHANGE: removes the `man`, `math`, `plan_agent` tools, the
plan-review SSE/HTTP/SDK surface, and the plan_* model-routing settings
from the experimental 1.6 line.
Replace linear optimization chain with UCB1 evolution tree. Each
iteration selects the most promising node to extend, preventing
irrecoverable collapse from bad edits. Also adds improvement-based
delta feedback to the optimizer and optional holdout set separation.
Inspired by "Learning to Self-Evolve" (arXiv:2603.18620).
Patrick Buckley