[FEAT]:- The solver agent with the harness - #11
Merged
Conversation
…nfig file and the models to become the solver agents
…the model orchestrator
…olber yaml (update)
… running only that validator against the nine skill folders; it does not execute Solve, contact GitHub/Pollinations, clone a repository, or run project tests.
elixpoo
added
FEATURE
….elixpo into feat/solve-pipeline
Collaborator
|
[CHANGE LOG] 📋 Changelog updated for PR #11. View changelog |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Changes Made
agents/solve/— a bounded solve squad:core.pyorchestrates fork resolution → comprehension → edit → verify → review;harness.pysupervises a CCR Node coding harness (qwen-coder) withRead/Glob/Grep/Edit/Writeonly, stripping all credentials from the harness environment;model.py/models.pyhandle strictly-validated structured plans, steps, and review verdicts;edit.pygrounds every edit to a retrieved in-repo file;git.pyruns allowlisted verification commands shell-free;failure.pyconverts failures into adoctor_pendinghandoff with cleanup manifest;verification_plan.pyinfers the missing verification command deterministically from manifests..github/workflows/solve.yml— runs Solve + Submit in one isolated runner after a successful Vet gate (or viaworkflow_dispatchowned-test mode), with a 20-minute wall clock and committed-backstate/solve.json/state/ledger.json.agents/submit/__main__.py— validates aready_to_submitsolve state, passes the PR title/body through theqwen-safetygate, pushes once, and opens a disclosed[ELIXPO]PR ending inFixes #N.agents/vet/— sharedestimated_minutes(1–15) gate,owned-testmode with writable-repository verification, andagents/vet/github.pymoved tolib/github/issues.pyfor reuse across squads.lib/solve_policy.py+config/solve.yaml— single declarative source for token, time, file, setup/verification-command, and blocked-change limits shared by Vet, Solve, and Submit.rtk/router.py—record_external_usageso harness-model tokens land in the ledger;lib/workspace.pyreworked to pass GitHub auth via process env never CLI args.skills/— 10 newSKILL.mdpackages (solve-bounded-issue,implement-exact-edit,verify-repository-change,submit-autonomous-pr, etc.), each withagents/openai.yamldiscovery metadata.agents/comprehend/bundle.py— decode/splat unpacking for the bounded repo bundle into the fork workspace.tests/—test_solve.py(397),test_submit.py,test_comprehend.py, plus vet/rtk/lib additions.config/models.yaml— solve roles toqwen-coder; docs (pollinations.md,refactor_plan.md,README.md) updated.Checklist
python -m pytest tests/clean.envchanges look SOPS-encrypted, not raw secrets