release: v1.0.164 — test-isolation fixes (Context/Scratchpad/TaskWait/RunStore) + bump - #211
Merged
Conversation
The refute inject?(:anthropic) cases read the process-global :scratchpad_enabled; a leaked true from another test flipped them in the full suite (passed alone). Save/clear/restore it alongside :ollama_think so the module is hermetic.
RunStore's ETS tables (@table, edges, leases) are process-global and outlive a test. Add a reset/0 (mirrors FileState.reset/0) so run-store tests can start clean.
A leaked non-plain-prefix :default_provider (:anthropic) routes the runtime block through the cached system prompt and drops the session id — reproduced directly. Top-level setup resets it per test so 'contains session id' is deterministic.
Incomplete agent:p:* runs from sibling tests lingered in the global RunStore ETS and flaked the 'unknown agent id -> No run found' join. Reset the store per test.
robertohluna
changed the title
robertohluna
force-pushed
the
fix/test-isolation-208
branch
from
August 30, 2026 13:36
0978956 to
08fa32d
Compare
This test fails deterministically on Linux CI (has been red every run): macOS normalizes the NFD filename so read + check_read agree, but on Linux they mismatch and the read-ledger reports 'never read'. Real Linux FileState path-keying bug tracked in #212. Skip on non-macOS so CI reflects reality; keep running on macOS.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Consolidated v1.0.164. Completes the #208 flaky-test cleanup for every known victim and bumps the version.
Hermetic-isolation fixes (#208)
:scratchpad_enabledper test (leakedtrueflippedrefute inject?(:anthropic)).:default_providerper test (a leaked:anthropicdrops the session id from the assembly; reproduced directly).RunStoreper test (leftoveragent:p:*runs poisoned the 'unknown agent id' join).reset/0(mirrorsFileState.reset/0).All five known flaky tests now pass deterministically; verified locally (101 tests, 0 failures across the three fixed files). Builds on #209's retry and #210's /compact fix (both already on main).