Repository navigation
[PILOT-06] Execute project POST checks with Smol model and trial capabilities - #677
Merged
Merged
Conversation
added 30 commits
October 23, 2026 16:45
# Conflicts: # backend/app/adapters/artifacts/__init__.py # backend/app/modules/checkers/api/external_policy.py # external_checkers/workstream-checker-service/src/main.rs
…to codex/terminal-systems-post-checkers # Conflicts: # .ci/behavior-ownership/partition.v1.json # backend/tests/test_ci_lane_catalogue.py
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Terminal Systems POST images need budgeted model calls, authorized predecessor material and isolated build/test execution. This PR adds the project image and the generic Smol capabilities it consumes. Project images own grading; the Rust service owns execution. Provider credentials, authorization and accounting remain in the trusted proxy.
Builds and candidate steps use fresh microVMs, pinned offline inputs, bounded resources and verified ART evidence. The superseded Docker/runsc trial implementation is removed. Migrations 0063 and 0064 follow main’s 0062; provider selection from #690 is incorporated.
Validation:
b2ce3fa45: two verified ART trial receipts with 17 steps each, four controlled model effects, no-call replay and confirmed database/storage/VM cleanup. Ten direct KVM cases separately cover predecessor custody and resource ceilings. Builder and candidate workloads apply the shared syscall filter.8012f2792passed caller disconnect, exact child VMM death and controller death/restart. Each produced no false completion or trial ART effects, cleaned owned resources and reused both admission slots. All 12 raw artifacts, JUnit results and independent cleanup receipts were rehashed and inspected. Earlier failed observer cases remain recorded as failures.6d2a3424d. The final review-ready Backend run 40419888844 passed all 27 jobs, includingtest. Its freshly downloaded and independently audited artifacts contain 11,416 unique completed tests, zero skips/deselections and all 17 database/storage cleanup receipts. The tested merge tree is identical to the reviewed head.Hosted wiring and the capped live OpenRouter call are tracked separately in stacked #692. These controlled-provider proofs do not establish an approved public revision journey, production Flow identity, substantive model-grading quality or the full Terminal Systems project. PRE model grants remain unsupported.
Exact scope and evidence:
.commitrail/changes/terminal-systems-project-checkers.md. Human merge is required; merge this PR before #692.