Skip to content

Implementation Plan: Produce the Complete Parametrized Conformance Test Class#3991

Merged
Trecek merged 2 commits into
developfrom
t4-p4-a2-wp1-produce-the-complete-parametrized-conformance-t/3936
Jun 10, 2026
Merged

Implementation Plan: Produce the Complete Parametrized Conformance Test Class#3991
Trecek merged 2 commits into
developfrom
t4-p4-a2-wp1-produce-the-complete-parametrized-conformance-t/3936

Conversation

@Trecek

@Trecek Trecek commented Jun 10, 2026

Copy link
Copy Markdown
Collaborator

Summary

Create tests/execution/backends/test_coding_agent_backend_conformance.py containing TestCodingAgentBackendConformance — a parametrized test class inheriting BackendContractBase, exercising the CodingAgentBackend Protocol contract across all registered backends (claude-code, codex). The class covers core properties (8 tests), sub-protocol factories (6 tests), command-builder contracts (6 tests), setup_session_dir branching (1 test), SessionLocator P3 methods (3 tests), capability-gated command builders (2 tests), and inspector not-capable assertion (1 test) — 27 test methods total.

Implementation Plan

Plan file: /home/talon/projects/autoskillit-runs/impl-20260610-120819-718947/.autoskillit/temp/make-plan/t4_p4_a2_wp1_conformance_test_plan_2026-06-10_121600.md

Closes #3936

🤖 Generated with Claude Code via AutoSkillit

Token Usage Summary

Step Model count uncached output cache_read peak_ctx turns cache_write time
plan* opus[1m] 1 1.6k 22.2k 1.7M 109.9k 55 88.7k 12m 43s
verify* sonnet 1 124 12.9k 784.5k 64.6k 38 43.6k 5m 31s
implement* MiniMax-M3 1 1.2M 5.8k 0 0 58 0 2m 47s
audit_impl* sonnet 1 44 14.6k 166.2k 41.8k 13 36.8k 4m 6s
prepare_pr* MiniMax-M3 1 203.2k 1.8k 0 0 12 0 47s
compose_pr* MiniMax-M3 1 213.6k 1.7k 0 0 15 0 40s
review_pr* sonnet 1 158 26.8k 1.1M 80.8k 47 60.9k 6m 11s
resolve_review* sonnet 1 163 10.3k 1.2M 72.2k 45 52.0k 6m 24s
Total 1.6M 96.1k 4.9M 109.9k 282.0k 39m 10s

* Step used a non-Anthropic provider; caching behavior may differ.

Token Efficiency

Step LoC Changed cache_read/LoC cache_write/LoC output/LoC
plan 0
verify 0
implement 195 0.0 0.0 29.7
audit_impl 0
prepare_pr 0
compose_pr 0
review_pr 0
resolve_review 3 386891.7 17349.0 3422.7
Total 198 24755.0 1424.3 485.6

Model Usage Breakdown

Model steps uncached output cache_read cache_write time
opus[1m] 1 1.6k 22.2k 1.7M 88.7k 12m 43s
sonnet 4 489 64.6k 3.2M 193.3k 22m 13s
MiniMax-M3 3 1.6M 9.3k 0 0 4m 13s

Trecek added 2 commits June 10, 2026 12:31
…class

Adds tests/execution/backends/test_coding_agent_backend_conformance.py
containing TestCodingAgentBackendConformance — a parametrized class
inheriting BackendContractBase that exercises the CodingAgentBackend
Protocol contract across all registered backends (claude-code, codex).
Covers core properties, sub-protocol factories, command-builder
contracts, setup_session_dir branching, SessionLocator P3 methods,
capability-gated command builders, and the inspector not-capable path
(27 test methods × 2 backends = 54 collected items).

Updates tests/execution/AGENTS.md backends table with the new file.
The pytest.fail branch is a forcing function that requires explicit
coverage for each new backend added to BACKEND_REGISTRY. A comment
now explains the design intent so future contributors add an elif
branch rather than removing the guard.
@Trecek Trecek added this pull request to the merge queue Jun 10, 2026
Merged via the queue into develop with commit 4cadec7 Jun 10, 2026
3 checks passed
@Trecek Trecek deleted the t4-p4-a2-wp1-produce-the-complete-parametrized-conformance-t/3936 branch June 10, 2026 20:06
Trecek added a commit that referenced this pull request Jun 28, 2026
…in WP2 disposition table

Updates §3 rows for #3273, #3385, and #2841 to include explicit references
to commit 4cadec7 (PR #3991, TestCodingAgentBackendConformance) and notes
on remaining B2 behavioral conformance gaps, satisfying Criterion 2.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
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.

1 participant