Skip to content

feat(examples): add 3 runtime failure demo packages - #158

Merged
dorado269 merged 3 commits into
dev.yihou.aiopt.task_packagefrom
dev.yaoc.aiopt.test_bad_cases
Sep 9, 2026
Merged

feat(examples): add 3 runtime failure demo packages#158
dorado269 merged 3 commits into
dev.yihou.aiopt.task_packagefrom
dev.yaoc.aiopt.test_bad_cases

Conversation

@ChengYao-amd

Copy link
Copy Markdown

No description provided.

@dorado269
dorado269 force-pushed the dev.yaoc.aiopt.test_bad_cases branch from dc8d493 to 292b8b3 Compare September 8, 2026 09:04
@dorado269
dorado269 changed the base branch from dev.yihou.aiopt.all.1 to dev.yaoc.aiopt.task_package September 8, 2026 09:05
ChengYao-amd and others added 3 commits September 8, 2026 09:19
Three deterministic, pure-program runtime failure packages that complement
the existing demo-reject-* load-time rejection family:

- demo-runtime-grounding-fail: pure-program version of demo's
  check_grounded FAIL path (produce → transform → consume chain)
- demo-runtime-fanout-partial: fan-out where one of three producers
  fails validation, starving the fan-in consumer
- demo-runtime-disagree: body succeeds but validator rejects due to
  content disagreement between two independent reviewers

All three:
- Use kind: program only (no AI, no credentials, no model calls)
- Pass stage B graph validation (--dry-run exit 0)
- Run end-to-end in ~30s with AGENT_SYS_NO_PERMISSIONS=1
- Exit 0 with 2/2 expected failures observed
- Have expectations.py entries for strict xfail accounting

Tested on smci355-ccs-aus-n05-29 (Linux 6.8.0, Python 3.12.14, Landlock).

Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
@dorado269
dorado269 force-pushed the dev.yaoc.aiopt.test_bad_cases branch from 292b8b3 to fa1bed5 Compare September 8, 2026 09:20
@dorado269
dorado269 changed the base branch from dev.yaoc.aiopt.task_package to dev.yihou.aiopt.task_package September 8, 2026 09:20
@dorado269
dorado269 merged commit 5fed156 into dev.yihou.aiopt.task_package Sep 9, 2026
1 of 2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants