Add fresh English delegation with verified local execution - #11
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add a standalone English handoff CLI: one fresh bounded local worker selects actions, generates raw Python source separately, writes/runs it, repairs from actual feedback, and returns observed results with reopenable raw evidence. Completion requires matching executed and saved bytes, exit zero and exact expected stdout; worker claims alone never pass.
Use non-quantized FP16 with the unchanged trained pilot adapter, rejecting quantized/unknown precision. Expose 32K context, 8K output, action and time budgets. Keep this prototype separate from the older file-inspection adapter; no new host tool registration or production change.
Verified on local Ollama through the public CLI: two fresh calculations, a real NameError repair, a final-revision fresh target, denial and action exhaustion. Reopened saved results/source/raw stdout. Compileall and all 13 core tests passed; CI adds the three delegation invariants. Earlier failed probes and limitations are recorded in ENGLISH_RESULTS.md.
Configured OpenCode2 delivery, host-native permissions, the actual PTY recorder, long-context quality and general speedup remain unverified. Generated Python has caller privileges; this is an explicitly authorized trusted local prototype, not a sandbox.