refactor: split opencode runtime and tests #17

2026-09-01T03:12:53Z

masi commented

2026-09-01 03:12:53 +00:00

Summary

reduce pilot/review/opencode.py to a 674-line compatibility and orchestration seam
split workspace preparation, lens configuration, lens execution, runtime, and synthesis into focused modules
split the 957-line opencode test monolith into workspace, lenses, and response test suites
update architecture and webhook documentation to match the current module layout
preserve compatibility exports and existing monkeypatch seams

Verification

python3 -m compileall -q pilot tests
python3 -m pytest tests/pilot -q
516 tests passed
git diff --check passed

## Summary - reduce pilot/review/opencode.py to a 674-line compatibility and orchestration seam - split workspace preparation, lens configuration, lens execution, runtime, and synthesis into focused modules - split the 957-line opencode test monolith into workspace, lenses, and response test suites - update architecture and webhook documentation to match the current module layout - preserve compatibility exports and existing monkeypatch seams ## Verification - python3 -m compileall -q pilot tests - python3 -m pytest tests/pilot -q - 516 tests passed - git diff --check passed

masi added 1 commit 2026-09-01 03:13:01 +00:00

refactor: split opencode runtime and tests f485d9faf9

pragent-bot bot reviewed 2026-09-01 03:20:30 +00:00

pragent-bot bot left a comment

🤖 AI Review · pragent pilot · headroom/MiniMax-M2.7 · f485d9fa · Merge confidence: 4/5 🟢

PR-Agent pilot on this repo. Comments are LLM-generated; treat as suggestions, not mandates.

Summary of Changes

Splits opencode.py workspace preparation into pilot/review/opencode_workspace.py (archive fetch, tar extraction, sanitization, brief rendering, factory drop)
Splits lens configuration into pilot/review/opencode_lens_config.py (ReviewerSpec dataclass, config parsing, reviewer resolution) — includes duplicate dead functions
Splits finding synthesis into pilot/review/opencode_synthesis.py (normalize, deduplicate, cap, tone-strip, summary field synthesis)
Splits lens execution into pilot/review/opencode_lenses.py and pilot/review/opencode_runtime.py (isolated subprocess, environment allow-listing, warm-up)

Key Risks & Concerns

Duplicate _coerce_str/_coerce_int definitions in opencode_lens_config.py shadow the first definitions with identical implementations — dead code and a merge artifact that should be removed
opencode.py imports many symbols from submodules that are unused in the module body and may be retained only for backward-compatibility re-exports — should be made explicit with all to avoid confusing future readers
opencode_synthesis.py imports concurrent.futures and dataclasses but never uses them — vestigial imports from the original monolith not cleaned up in the refactor

Findings Overview

1 inline comment(s); 1 total.

Severity	Location	Finding
🟡 [MEDIUM]	`pilot/review/opencode_lens_config.py:58`	_coerce_str is defined twice in this module — lines 29-30 and lines 58-59. The second definition shadows the first with an identical implementation. Same applies to _coerce_int at lines 33-38 and 62-67. This is a merge artifact that leaves dead code and could confuse future maintainers about which definition is authoritative.

🔋 AI Usage & Run Details

Model / Engine: headroom/MiniMax-M2.7 · opencode · 14 steps · 442.0s
Total Tokens: 147,095 (147.1K) in / 9,003 (9.0K) out (0 reasoning, cache 720,285 (720.3K) read / 72,391 (72.4K) write, 948,774 (948.8K) total)
Actual: $0.00 (headroom/MiniMax-M2.7 — free tier)
Scope: Whole-repo checkout at head sha (agent can read any file + run linters, not just the diff) — input tokens include files read beyond the diff. Per-comment output is attributed (one model pass produces all findings; output split by each finding's body weight).

🤖 **AI Review** · pragent pilot · headroom/MiniMax-M2.7 · `f485d9fa` · Merge confidence: 4/5 🟢 > PR-Agent pilot on this repo. Comments are LLM-generated; treat as suggestions, not mandates. ### Summary of Changes - Splits opencode.py workspace preparation into pilot/review/opencode_workspace.py (archive fetch, tar extraction, sanitization, brief rendering, factory drop) - Splits lens configuration into pilot/review/opencode_lens_config.py (ReviewerSpec dataclass, config parsing, reviewer resolution) — includes duplicate dead functions - Splits finding synthesis into pilot/review/opencode_synthesis.py (normalize, deduplicate, cap, tone-strip, summary field synthesis) - Splits lens execution into pilot/review/opencode_lenses.py and pilot/review/opencode_runtime.py (isolated subprocess, environment allow-listing, warm-up) ### Key Risks & Concerns - Duplicate _coerce_str/_coerce_int definitions in opencode_lens_config.py shadow the first definitions with identical implementations — dead code and a merge artifact that should be removed - opencode.py imports many symbols from submodules that are unused in the module body and may be retained only for backward-compatibility re-exports — should be made explicit with __all__ to avoid confusing future readers - opencode_synthesis.py imports concurrent.futures and dataclasses but never uses them — vestigial imports from the original monolith not cleaned up in the refactor ### Findings Overview _1 inline comment(s); 1 total._ | Severity | Location | Finding | |---|---|---| | 🟡 [MEDIUM] | `pilot/review/opencode_lens_config.py:58` | _coerce_str is defined twice in this module — lines 29-30 and lines 58-59. The second definition shadows the first with an identical implementation. Same applies to _coerce_int at lines 33-38 and 62-67. This is a merge artifact that leaves dead code and could confuse future maintainers about which definition is authoritative. | <details> <summary>🔋 AI Usage & Run Details</summary> - **Model / Engine**: `headroom/MiniMax-M2.7` · opencode · 14 steps · 442.0s - **Total Tokens**: 147,095 (147.1K) in / 9,003 (9.0K) out (0 reasoning, cache 720,285 (720.3K) read / 72,391 (72.4K) write, 948,774 (948.8K) total) - **Actual**: $0.00 (headroom/MiniMax-M2.7 — free tier) - **Scope**: Whole-repo checkout at head sha (agent can read any file + run linters, not just the diff) — input tokens include files read beyond the diff. Per-comment output is *attributed* (one model pass produces all findings; output split by each finding's body weight). </details>

pilot/review/opencode_lens_config.py

						
				@@ -0,0 +55,4 @@

				    ]

				def _coerce_str(v, default: str = "") -> str:

pragent-bot bot commented

2026-09-01 03:20:30 +00:00

🟡 [MEDIUM] _coerce_str is defined twice in this module — lines 29-30 and lines 58-59. The second definition shadows the first with an identical implementation. Same applies to _coerce_int at lines 33-38 and 62-67. This is a merge artifact that leaves dead code and could confuse future maintainers about which definition is authoritative.

Fix: Remove the duplicate definitions at lines 58-67; keep the first definitions at lines 29-38.

def _coerce_str(v, default: str = "") -> str:
    return str(v).strip() if isinstance(v, (str, int, float)) else default


def _coerce_int(v, default: int, lo: int, hi: int) -> int:
    try:
        n = int(v)
    except (TypeError, ValueError):
        return default
    return max(lo, min(hi, n))

🪙 ~9,003 (9.0K) tok (100% · attributed output)

🟡 [MEDIUM] _coerce_str is defined twice in this module — lines 29-30 and lines 58-59. The second definition shadows the first with an identical implementation. Same applies to _coerce_int at lines 33-38 and 62-67. This is a merge artifact that leaves dead code and could confuse future maintainers about which definition is authoritative. **Fix:** Remove the duplicate definitions at lines 58-67; keep the first definitions at lines 29-38. ```suggestion def _coerce_str(v, default: str = "") -> str: return str(v).strip() if isinstance(v, (str, int, float)) else default def _coerce_int(v, default: int, lo: int, hi: int) -> int: try: n = int(v) except (TypeError, ValueError): return default return max(lo, min(hi, n)) ``` 🪙 ~9,003 (9.0K) tok (100% · attributed output)

masi marked this conversation as resolved

masi added 1 commit 2026-09-01 03:27:55 +00:00

fix: remove duplicate lens config helpers e6cb7d01a6

pragent-bot bot reviewed 2026-09-01 03:33:44 +00:00

pragent-bot bot left a comment

🤖 AI Review · pragent pilot · headroom/MiniMax-M2.7 · e6cb7d01 · Merge confidence: 5/5 🟢

PR-Agent pilot on this repo. Comments are LLM-generated; treat as suggestions, not mandates.

Summary of Changes

Split pilot/review/opencode.py (957 lines) into opencode_workspace.py, opencode_lens_config.py, opencode_lenses.py, opencode_synthesis.py, and opencode_runtime.py — preserving all exports via a compatibility seam
Added backward-compatibility shim at pilot/opencode_review.py that redirects to the new module path
Split the opencode test monolith into opencode_lenses_test.py and opencode_response_test.py focused suites
Updated architecture.md, pilot/README-webhook.md, and pilot/README.md to reflect the new module layout

Key Risks & Concerns

None identified.

🔋 AI Usage & Run Details

Model / Engine: headroom/MiniMax-M2.7 · opencode · 12 steps · 328.5s
Total Tokens: 46,039 (46.0K) in / 6,158 (6.2K) out (0 reasoning, cache 461,186 (461.2K) read / 57,269 (57.3K) write, 570,652 (570.7K) total)
Actual: $0.00 (headroom/MiniMax-M2.7 — free tier)
Scope: Whole-repo checkout at head sha (agent can read any file + run linters, not just the diff) — input tokens include files read beyond the diff. Per-comment output is attributed (one model pass produces all findings; output split by each finding's body weight).

🤖 **AI Review** · pragent pilot · headroom/MiniMax-M2.7 · `e6cb7d01` · Merge confidence: 5/5 🟢 > PR-Agent pilot on this repo. Comments are LLM-generated; treat as suggestions, not mandates. ### Summary of Changes - Split pilot/review/opencode.py (957 lines) into opencode_workspace.py, opencode_lens_config.py, opencode_lenses.py, opencode_synthesis.py, and opencode_runtime.py — preserving all exports via a compatibility seam - Added backward-compatibility shim at pilot/opencode_review.py that redirects to the new module path - Split the opencode test monolith into opencode_lenses_test.py and opencode_response_test.py focused suites - Updated architecture.md, pilot/README-webhook.md, and pilot/README.md to reflect the new module layout ### Key Risks & Concerns _None identified._ <details> <summary>🔋 AI Usage & Run Details</summary> - **Model / Engine**: `headroom/MiniMax-M2.7` · opencode · 12 steps · 328.5s - **Total Tokens**: 46,039 (46.0K) in / 6,158 (6.2K) out (0 reasoning, cache 461,186 (461.2K) read / 57,269 (57.3K) write, 570,652 (570.7K) total) - **Actual**: $0.00 (headroom/MiniMax-M2.7 — free tier) - **Scope**: Whole-repo checkout at head sha (agent can read any file + run linters, not just the diff) — input tokens include files read beyond the diff. Per-comment output is *attributed* (one model pass produces all findings; output split by each finding's body weight). </details>

masi merged commit a202bd3598 into main

2026-09-01 03:35:19 +00:00

masi deleted branch refactor/split-opencode-modules

2026-09-01 03:35:21 +00:00

Sign in to join this conversation.

2 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: gitea_admin/pragent#17