Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 14 additions & 11 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -203,13 +203,14 @@ configured promotion policy.
Training rounds run through the provider-neutral `TrainingLoop` contract.
This package registers `createHarnessLoop()` as the default, so
`ts-autocode-harness` owns bounded rounds, feedback, cancellation, and stall
detection: the same callback path accepts arbitrary judge inputs, requires an
exact pass/fail decision, tests
approved candidates with an isolated adversary, and makes the teacher revise
the rubric when the adversary exposes an accepted gap. Baseline results are
never treated as proof that a rewrite passes. Set `TrainingSettings.loop` to
substitute your own orchestration; the lower-level `evaluate` and
`ts-autocode-rewrite` promotion primitives also remain available.
detection. By default, training reviews serve as the harness's evidence — a
configured judge may decide differently: a candidate
passes exactly when its review reports no gate failures, accepted candidates
are re-reviewed by an isolated adversary, and a standing challenge tightens
the rubric before the next round. Baseline results are never treated as proof
that a rewrite passes. Set `TrainingSettings.loop` to substitute your own
orchestration; the lower-level `evaluate` and `ts-autocode-rewrite`
promotion primitives also remain available.

The built-in loop is an observable round sequence (`trainingRounds()`
pushes each reviewed round to a subscriber; `sequentialLoop` collects the
Expand All @@ -235,10 +236,12 @@ Runtime dependencies enter through `TrainingSettings`:
is not the desired project.
- `outputDir` relocates run artifacts and eval output (default `.agentv`,
exported as `defaultOutputDir`); a run's `EvalConfig.outputDir` still
overrides it. `createHarnessLoop({ actionLogFile, contextProvider })`
renames the write-ahead action log inside that directory and replaces the
default rolling-window context management (`windowedContext`) with your own
provider — for example a rolling-summary reducer.
overrides it. Every `createHarnessLoop` collaborator is injectable:
`bus` builds the run's message bus (replacing the default file-backed
write-ahead log named by `actionLogFile` inside that directory), `judge`
gates every harness action and verdict, and `contextProvider` replaces the
default rolling-window context management (`windowedContext`) — for example
with a rolling-summary reducer.

AgentV's `workers` option parallelizes live-trace and candidate evals. Independent
trainables can be trained concurrently by the application, while the configured
Expand Down
Loading
Loading