Skip to content

Tests: slim test_orchestrator.py to behaviour-focused cases (#497 step 1) - #507

Merged
renmengye merged 7 commits into
mainfrom
tests/slim-orchestrator
Oct 11, 2026
Merged

renmengye merged 7 commits into
mainfrom
tests/slim-orchestrator

Conversation

@renmengye

Copy link
Copy Markdown
Member

Step 1 of #497 (fewer, behaviour-focused tests). Tests only; src/ is unchanged.

tests/test_orchestrator.py:

before after
tests 250 (+4 skipped) 179
runtime 4.9 s 1.2 s
covered lines of outerloop.orchestrator 736/863 736/863, the same line set
length 3,370 lines about 3,190 lines
  • The scope-refusal matrix (4 kinds × 7 retries × 3 entries = 84 cases) now has 11 cases, one per distinct path: the three entry words, replies-only wiring, refusal into a resume entry, the author's answers (end, and error, budget and outage wakes), and recovery.
  • 18 hand-rolled attempt_once call sites go through the existing run_climb helper, so a change to the call is fixed in one place.
  • Four tests that duplicated a covered path were deleted, and one incidental call-count pin was dropped. Exact call-sequence and text assertions stay where they are the test's subject.
  • Codex review found that two behaviours had lost their outcome assertions while line coverage stayed equal: run-level crediting at an exact decimal floor and refusal under an infinite floor, and the negative-result note of a minimize-direction regression. Both are restored as assertions in existing run-level tests (50dfeda).

Built by GLM-5.3 (self-reviewed, coverage-gated), reviewed by Codex, final read by Claude Opus 5.5.

@github-actions github-actions Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Round 1 — reviewed head 50dfedac — reviewer summarizer:hermes/gpt-5.6-terra over coverage+credentials+deployment+general+lifecycle+prose.

terra
Advisory findings from outerloop — the code owner decides. Reply to disagree; the outerloop:no-review label opts this PR out.

Verdict: nothing blocking — 2 advisory notes.

2 findings attached to the lines below.

Two non-blocking test-coverage regressions: the report-less submit case no longer enforces a single author session, and the scope-refusal matrix no longer covers timeout budget classification. Rejected findings: none.

Comment thread tests/test_orchestrator.py Outdated
Comment thread tests/test_orchestrator.py
@renmengye renmengye added the outerloop:review re-request the advisory review label Oct 11, 2026

@github-actions github-actions Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Round 2 — reviewed head df77a6c0 — reviewer hermes/gpt-5.6-terra.

terra
Advisory findings from outerloop — the code owner decides. Reply to disagree; the outerloop:no-review label opts this PR out.

Verdict: no defects found.

@renmengye
renmengye merged commit 8901fc2 into main Oct 11, 2026
5 checks passed
@renmengye
renmengye deleted the tests/slim-orchestrator branch October 11, 2026 07:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

outerloop:review re-request the advisory review

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant