Skip to content

Implementation Plan: T5-P3-A2-WP1 Behavioral Conformance Suite Extension#4081

Merged
Trecek merged 2 commits into
developfrom
t5-p3-a2-wp1-extend-the-existing-parametrized-conformance-cl/4009
Jun 12, 2026
Merged

Implementation Plan: T5-P3-A2-WP1 Behavioral Conformance Suite Extension#4081
Trecek merged 2 commits into
developfrom
t5-p3-a2-wp1-extend-the-existing-parametrized-conformance-cl/4009

Conversation

@Trecek

@Trecek Trecek commented Jun 12, 2026

Copy link
Copy Markdown
Collaborator

Summary

Extend tests/execution/backends/test_coding_agent_backend_conformance.py with Group 6 behavioral tests that exercise three BackendCapabilities fields (mcp_env_forward_vars, session_resume_capable, setup/locator round-trip), plus two module-level meta-tests and two module-level registry constants (NOT_YET_LIVE, CAPABILITY_CLASSIFICATION) that guard against new capability fields being added without conformance coverage or classification. Also add docstrings to all existing Group 1–5 test methods citing the BackendCapabilities field(s) each exercises, which is required for the test_every_capability_field_exercised_or_not_yet_live meta-test to pass.

Implementation Plan

Plan file: /home/talon/projects/autoskillit-runs/impl-20260611-194852-163666/.autoskillit/temp/make-plan/t5-p3-a2-wp1-behavioral-conformance-suite_plan_2026-06-11_195600.md

Closes #4009

🤖 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 2.0k 33.2k 1.4M 116.4k 43 102.5k 15m 16s
verify* sonnet 1 2.6k 14.7k 572.7k 68.3k 31 47.3k 7m 41s
implement* MiniMax-M3 1 4.2M 20.8k 0 0 103 0 8m 51s
audit_impl* sonnet 1 44 18.3k 168.9k 44.5k 15 43.0k 6m 6s
prepare_pr* MiniMax-M3 1 240.1k 2.3k 0 0 14 0 1m 1s
compose_pr* MiniMax-M3 1 175.7k 1.1k 0 0 11 0 40s
review_pr* sonnet 1 156 26.9k 1.1M 77.2k 46 56.6k 6m 49s
resolve_review* sonnet 1 187 13.2k 1.5M 84.6k 58 64.3k 7m 31s
Total 4.6M 130.6k 4.8M 116.4k 313.8k 53m 57s

* 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 183 0.0 0.0 113.5
audit_impl 0
prepare_pr 0
compose_pr 0
review_pr 0
resolve_review 6 254465.0 10720.3 2198.0
Total 189 25246.7 1660.2 690.9

Model Usage Breakdown

Model steps uncached output cache_read cache_write time
opus[1m] 1 2.0k 33.2k 1.4M 102.5k 15m 16s
sonnet 4 3.0k 73.2k 3.4M 211.2k 28m 7s
MiniMax-M3 3 4.6M 24.2k 0 0 10m 32s

Trecek and others added 2 commits June 11, 2026 20:20
- Add Group 6 parametrized behavioral tests covering mcp_env_forward_vars,
  session_resume_capable, and setup_session_dir→session_locator round-trip
- Add module-level constants NOT_YET_LIVE (7 fields) and
  CAPABILITY_CLASSIFICATION (41 REQUIRED/OPTIONAL fields)
- Add module-level meta-tests: every field exercised or NOT_YET_LIVE,
  every field has a REQUIRED/OPTIONAL classification
- Add docstrings to existing Group 1-5 test methods citing the
  BackendCapabilities field(s) each exercises

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…st_every_capability_field_has_classification

Closes the cross-validation gap between NOT_YET_LIVE and CAPABILITY_CLASSIFICATION:
a field reclassified from OPTIONAL to REQUIRED while still in NOT_YET_LIVE would
silently skip coverage enforcement for a required capability.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@Trecek Trecek added this pull request to the merge queue Jun 12, 2026
Merged via the queue into develop with commit 56ee09a Jun 12, 2026
3 checks passed
@Trecek Trecek deleted the t5-p3-a2-wp1-extend-the-existing-parametrized-conformance-cl/4009 branch June 12, 2026 03:57
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