feat(email-writing): compose bounded judged review service - #1535
Conversation
|
Important Draft PR not reviewedDraft PRs are not automatically reviewed by default.
To automatically review draft PRs, update your CodeRabbit configuration: reviews:
auto_review:
drafts: trueThanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
@coderabbitai review Please review the exact current Task-9 head |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
set -euo pipefail
head='9670ba049ae23e9396051d21ead9194197de55eb'
base='21fa293d967bac0dd2b7b9a917b9fbce415134a2'
echo '== HEAD AND ANCESTRY =='
git rev-parse HEAD
git show -s --format='commit=%H%nparents=%P%nsubject=%s' "$head"
git merge-base --is-ancestor "$base" "$head" && echo "base_is_ancestor=yes" || echo "base_is_ancestor=no"
echo '== TASK-9 CHANGED PATHS =='
git diff --name-status "$base" "$head"
echo '== CANDIDATE FILES =='
fd -t f . | rg -i '(email.*writing|review.*service|candidate|judge|evidence|policy|workflow)'
echo '== PYTHON OUTLINES FOR CHANGED FILES =='
git diff --name-only "$base" "$head" -- '*.py' | while IFS= read -r f; do
echo "--- $f"
ast-grep outline "$f" --items all --type class,function --view expanded || true
done
echo '== CONTRACT-RELATED REFERENCES IN CHANGED PYTHON =='
git diff --name-only "$base" "$head" -- '*.py' | while IFS= read -r f; do
echo "--- $f"
rg -n -C 4 -i 'run_judge|to_thread|candidate|judge|policy|persist|evidence|cancel|abstain|diagnostic|send|publish|regex|keyword|language|timeout|deadline|revision|selector' "$f" || true
done
echo '== PR CHECKS FOR REQUESTED HEAD =='
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 82463 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== REVIEW SERVICE: COMPLETE NUMBERED SOURCE =='
cat -n backend/services/email_writing_review_service.py
echo '== REVIEW CONTRACTS OUTLINE AND REVIEW TYPES =='
ast-grep outline backend/services/email_writing_contracts.py --items all --type class,function --view expanded
rg -n -C 8 'class EmailWritingReview(Request|Response)|class EmailWritingDiagnostic|class EmailWritingDocumentGuidance|ReviewStatus|class EmailWritingProvenance' backend/services/email_writing_contracts.py
echo '== POLICY ADMISSION IMPLEMENTATION =='
ast-grep outline backend/services/email_writing_policy.py --items all --type class,function --view expanded
rg -n -C 8 'def evaluate_policy_admission|type AdmissionOutcome|AdmissionOutcome|publish_decision|evaluation_only|withhold|admit' backend/services/email_writing_policy.py backend/policies/email_writing_judge_evaluation_only_v1.json
echo '== EVIDENCE ORM AND MIGRATION =='
ast-grep outline backend/db/email_writing_evidence.py --items all --type class,function --view expanded
cat -n backend/db/email_writing_evidence.py
cat -n backend/alembic/versions/20260812_0001_add_email_writing_review_evidence.py
echo '== TASK-9 WORKFLOW =='
cat -n .github/workflows/email-writing-review-service-tdd.yml
echo '== IMPORTS OR USES OF THE TASK-9 SERVICE =='
rg -n -C 4 'EmailWritingReviewService|email_writing_review_service' backend --glob '*.py'
echo '== DIFF WITH CONTEXT =='
git diff --find-renames --unified=80 21fa293d967bac0dd2b7b9a917b9fbce415134a2 9670ba049ae23e9396051d21ead9194197de55eb -- \
backend/services/email_writing_review_service.py \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
.github/workflows/email-writing-review-service-tdd.ymlLength of output: 215890 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== CANDIDATE REVIEW: OUTLINE AND VALIDATION PATHS =='
ast-grep outline backend/services/email_writing_candidate_review.py --items all --type class,function --view expanded
rg -n -C 8 'class EmailWritingCandidate(ReviewResult|Diagnostic|Output)|def .*valid|selector|draft|revision|projection|language|evidence|hash|orchestration' backend/services/email_writing_candidate_review.py
echo '== INDEPENDENT JUDGE: OUTLINE AND EVALUATION PATHS =='
ast-grep outline backend/services/email_writing_judge.py --items all --type class,function --view expanded
rg -n -C 8 'class EmailWritingIndependentJudge|def evaluate|candidate|provider|fast_mlsirm|to_thread|executor|send|publish|admission' backend/services/email_writing_judge.py
echo '== TASK-5 BOUNDED EXECUTOR CONTRACTS =='
rg -n -C 8 'def run_judge|run_judge\s*=|class .*Judge.*(Executor|Worker)|JudgeExecutor|bounded.*judge|judge.*bounded' backend --glob '*.py'
echo '== TIMEOUT AND PERSISTENCE CONTROL FLOW =='
sed -n '332,409p' backend/services/email_writing_review_service.py
sed -n '728,935p' backend/services/email_writing_review_service.py
rg -n -C 6 'review_timeout|total_wall|timeout' backend/tests --glob '*email_writing*.py'
echo '== WORKFLOW/PATH AUTHORITY SEARCH =='
rg -n -i -C 3 'email.writing.review|email_writing_review|EmailWritingReviewService|send.*email|publish.*diagnostic|review.*api' backend/api backend/services --glob '*.py'Length of output: 114260
I found two issues.
The remaining requested boundaries are present in the reviewed source:
You are interacting with an AI system. |
|
@coderabbitai review Please re-review exact current Task-9 head |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
set -euo pipefail
head='163e943358dcaac53f978064fe5a7a30a7a17423'
base='21fa293d967bac0dd2b7b9a917b9fbce415134a2'
echo '== EXACT HEAD =='
git rev-parse HEAD
git show -s --format='commit=%H%nparents=%P%nsubject=%s' "$head"
git merge-base --is-ancestor "$base" "$head" && echo 'base_is_ancestor=yes' || echo 'base_is_ancestor=no'
git diff --name-status "$base" "$head"
echo '== SERVICE OUTLINE =='
ast-grep outline backend/services/email_writing_review_service.py --items all --type class,function --view expanded
echo '== SERVICE: CANDIDATE VALIDATION, DEADLINES, CANCELLATION, AND PERSISTENCE =='
rg -n -C 12 -i \
'validate_candidate|validate.*selector|validate.*evidence|run_judge|asyncio\.timeout|TimeoutError|CancelledError|post.*timeout|persist|commit|rollback|review_evidence_unavailable|finalize_non_admitted|candidate_result' \
backend/services/email_writing_review_service.py
echo '== PRECISE SERVICE RANGES =='
sed -n '1,460p' backend/services/email_writing_review_service.py
sed -n '460,1040p' backend/services/email_writing_review_service.py
echo '== RELATED CONTRACTS AND VALIDATOR IMPLEMENTATIONS =='
ast-grep outline backend/services/email_writing_candidate_review.py --items all --type class,function --view expanded
ast-grep outline backend/services/email_writing_contracts.py --items all --type class,function --view expanded
rg -n -C 10 -i \
'validate.*selector|validate.*evidence|selector|evidence_id|current_draft|authorized|candidate.*port|candidate.*result' \
backend/services/email_writing_candidate_review.py backend/services/email_writing_contracts.py
echo '== FOCUSED REGRESSION TESTS =='
for f in backend/tests/test_email_writing_review_service.py backend/tests/test_email_writing_review_service_execution.py; do
echo "--- $f"
ast-grep outline "$f" --items all --type class,function --view expanded
rg -n -C 10 -i \
'out.of.range|unknown.*evidence|selector|injected.*candidate|timeout|commit|rollback|cancel|review_evidence_unavailable|run_judge' "$f"
done
echo '== CURRENT-HEAD DIFF FOR REPAIRS =='
git diff --find-renames --unified=100 \
9670ba049ae23e9396051d21ead9194197de55eb "$head" -- \
backend/services/email_writing_review_service.py \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
.github/workflows/email-writing-review-service-tdd.yml
echo '== RUNTIME AUTHORITY AND EXECUTOR SEARCH =='
rg -n -C 5 -i \
'EmailWritingReviewService|email_writing_review_service|asyncio\.to_thread|run_judge|send.*email|publish.*diagnostic|review.*api' \
backend --glob '*.py'
echo '== TASK-9 WORKFLOW AND LIVE PR CHECKS =='
cat -n .github/workflows/email-writing-review-service-tdd.yml
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 173845 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== EXACT REPAIR SOURCE =='
nl -ba backend/services/email_writing_review_service.py | sed -n '240,430p'
nl -ba backend/services/email_writing_review_service.py | sed -n '800,970p'
echo '== SHARED CANDIDATE VALIDATOR =='
rg -n -A120 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_review_service.py backend/services/email_writing_candidate_review.py
echo '== COMPLETE TASK-9 REGRESSION TEST FILE =='
nl -ba backend/tests/test_email_writing_review_service_review_regressions.py
echo '== FOCUSED TEST NAMES AND ASSERTIONS =='
rg -n -A40 -B5 \
'^async def test_|^def test_|block_commit|block_rollback|rollbacks|commits|review_evidence_unavailable|candidate_selector_out_of_range|candidate_evidence_unknown|Judge must not run' \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py
echo '== ASYNCIO TIMEOUT IMPLEMENTATION AVAILABLE TO THE ENVIRONMENT =='
python - <<'PY'
import asyncio
import inspect
print(inspect.getsource(asyncio.Timeout))
PYLength of output: 1583 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== EXACT REPAIR SOURCE: VALIDATION AND TIMEOUT =='
sed -n '240,430p' backend/services/email_writing_review_service.py | cat -n
echo '== EXACT REPAIR SOURCE: FINALIZATION AND PERSISTENCE =='
sed -n '700,980p' backend/services/email_writing_review_service.py | cat -n
echo '== SHARED CANDIDATE VALIDATOR =='
rg -n -A120 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_review_service.py backend/services/email_writing_candidate_review.py
echo '== COMPLETE TASK-9 REGRESSION TEST FILE =='
cat -n backend/tests/test_email_writing_review_service_review_regressions.py
echo '== FOCUSED TESTS FOR THE REPAIRS =='
rg -n -A55 -B5 \
'^async def test_|^def test_|block_commit|block_rollback|rollbacks|commits|review_evidence_unavailable|candidate_selector_out_of_range|candidate_evidence_unknown|Judge must not run' \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.pyLength of output: 110790
I found one remaining issue.
The Candidate repair is correct in the reviewed source:
The service still uses the injected You are interacting with an AI system. |
seonghobae
added
enhancement
ChatGPT Codex Connector
|
@coderabbitai review Please re-review exact current Task-9 head |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
set -euo pipefail
head='88cffb54a981a3d7699e524947f81d7659bd71ad'
previous='163e943358dcaac53f978064fe5a7a30a7a17423'
base='21fa293d967bac0dd2b7b9a917b9fbce415134a2'
echo '== HEAD, ANCESTRY, AND CHANGED PATHS =='
git rev-parse HEAD
git show -s --format='commit=%H%nparents=%P%nsubject=%s' "$head"
git merge-base --is-ancestor "$base" "$head" && echo 'base_is_ancestor=yes' || echo 'base_is_ancestor=no'
git diff --name-status "$previous" "$head"
echo '== REVIEW SERVICE OUTLINE =='
ast-grep outline backend/services/email_writing_review_service.py --items all --type class,function --view expanded
echo '== REVIEW SERVICE: DEADLINES, CLEANUP, PERSISTENCE, CANCELLATION =='
rg -n -C 14 -i \
'ROLLBACK_SECONDS|rollback|commit|persist_review|evidence_seconds|asyncio\.timeout|TimeoutError|CancelledError|shield|create_task|review_evidence_unavailable|finalize_non_admitted|finalize.*cancel|cancel' \
backend/services/email_writing_review_service.py
echo '== REVIEW SERVICE: CANDIDATE VALIDATION AND JUDGE ORDER =='
rg -n -C 12 \
'_validate_candidate_result_binding|_validate_candidate_selectors_and_evidence|candidate_reviewer|run_judge|evaluate_policy|diagnostic' \
backend/services/email_writing_review_service.py
echo '== PRECISE SERVICE RANGES =='
nl -ba backend/services/email_writing_review_service.py | sed -n '1,470p'
nl -ba backend/services/email_writing_review_service.py | sed -n '700,1040p'
echo '== CANDIDATE SHARED VALIDATOR =='
rg -n -A140 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_candidate_review.py \
backend/services/email_writing_review_service.py
echo '== FOCUSED TEST OUTLINES AND TIMEOUT/CANCELLATION ASSERTIONS =='
for f in \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py
do
echo "--- $f"
ast-grep outline "$f" --items all --type class,function --view expanded
rg -n -C 10 -i \
'blocked.*commit|blocked.*rollback|rollback|commit|timeout|deadline|cancel|review_evidence_unavailable|out.of.range|unknown.*evidence|Judge must not run|run_judge' \
"$f"
done
echo '== REPAIR DIFF =='
git diff --find-renames --unified=100 "$previous" "$head" -- \
backend/services/email_writing_review_service.py \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py \
.github/workflows/email-writing-review-service-tdd.yml
echo '== RUNTIME AUTHORITY SEARCH =='
rg -n -C 4 -i \
'EmailWritingReviewService|email_writing_review_service|asyncio\.to_thread|run_judge|send.*email|publish.*diagnostic|review.*api' \
backend --glob '*.py'
echo '== LIVE PR CHECKS =='
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 63982 🏁 Script executed: #!/bin/bash
set -euo pipefail
head='88cffb54a981a3d7699e524947f81d7659bd71ad'
previous='163e943358dcaac53f978064fe5a7a30a7a17423'
echo '== COMPLETE REVIEW SERVICE WITH LINE NUMBERS =='
cat -n backend/services/email_writing_review_service.py
echo '== COMPLETE REGRESSION TEST FILE WITH LINE NUMBERS =='
cat -n backend/tests/test_email_writing_review_service_review_regressions.py
echo '== CANCELLATION TESTS IN ALL FOCUSED TASK-9 TESTS =='
rg -n -A90 -B12 \
'CancelledError|review_cancelled|cancel\(|shield\(|persist_terminal_review|rollback' \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py
echo '== TIMEOUT AND BLOCKED PERSISTENCE TESTS =='
rg -n -A110 -B12 \
'block_commit|block_rollback|commit_started|rollback_started|review_evidence_unavailable|TIMEOUT_EVIDENCE|ROLLBACK_SECONDS|total_wall_seconds' \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py
echo '== EXACT REPAIR DIFF =='
git diff --find-renames --unified=120 "$previous" "$head" -- \
backend/services/email_writing_review_service.py \
backend/tests/test_email_writing_review_service_review_regressions.py
echo '== WORKFLOW FOCUSED TEST COMMANDS =='
cat -n .github/workflows/email-writing-review-service-tdd.yml
echo '== EXACT PR CHECK STATE =='
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || true
echo '== CHECKOUT CONTENT ID =='
git rev-parse HEAD
git diff --quiet "$head" -- && echo 'checkout_matches_requested_head=yes' || echo 'checkout_matches_requested_head=no'Length of output: 171996
I found no new actionable defects. The rollback repair resolves the prior finding. The regression test covers both cases:
The test requires The Candidate trust boundary remains correct. The non-admitted paths return no semantic diagnostics or document guidance. The service builds user-facing diagnostics only after independent Judge execution, policy admission, confidence validation, and successful evidence persistence. The inspected service source does not send email or mutate email state. The checkout matches the requested head. The Task-9 predecessor is an ancestor of this head. Workflow run You are interacting with an AI system. |
|
@coderabbitai review Please review the exact current Task-9 head |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
set -euo pipefail
head='88cffb54a981a3d7699e524947f81d7659bd71ad'
base='21fa293d967bac0dd2b7b9a917b9fbce415134a2'
echo '== EXACT COMMIT AND ANCESTRY =='
echo "checkout=$(git rev-parse HEAD)"
git show -s --format='commit=%H%nparents=%P%nsubject=%s' "$head"
git merge-base --is-ancestor "$base" "$head" && echo 'task8_predecessor_is_ancestor=yes' || echo 'task8_predecessor_is_ancestor=no'
git diff --quiet "$head" -- && echo 'checkout_matches_requested_head=yes' || echo 'checkout_matches_requested_head=no'
echo '== CHANGED PATHS FROM TASK-8 PREDECESSOR =='
git diff --name-status "$base" "$head"
echo '== REVIEW SERVICE OUTLINE =='
ast-grep outline backend/services/email_writing_review_service.py --items all --type class,function --view expanded
echo '== REVIEW SERVICE: ORDER, VALIDATION, JUDGE, POLICY, PERSISTENCE, CANCELLATION =='
nl -ba backend/services/email_writing_review_service.py | sed -n '1,470p'
nl -ba backend/services/email_writing_review_service.py | sed -n '700,1060p'
echo '== SHARED CANDIDATE VALIDATOR =='
rg -n -A150 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_candidate_review.py \
backend/services/email_writing_review_service.py
echo '== JUDGE PORT AND POLICY =='
rg -n -C8 \
'run_judge|asyncio\.to_thread|ThreadPoolExecutor|evaluate_policy_admission|evaluation_only|withhold|admit' \
backend/services/email_writing_judge.py \
backend/services/email_writing_policy.py \
backend/policies/email_writing_judge_evaluation_only_v1.json \
backend/services/email_writing_review_service.py
echo '== EVIDENCE MODEL AND PERSISTED-FIELD REFERENCES =='
ast-grep outline backend/db/email_writing_evidence.py --items all --type class,function --view expanded
cat -n backend/db/email_writing_evidence.py
rg -n -C6 \
'EmailWritingReviewService|email_writing_review_service|send_email|send.*email|publish.*diagnostic|review.*api' \
backend --glob '*.py'
echo '== FOCUSED REGRESSIONS: TIMEOUT, ROLLBACK, CANCELLATION, CANDIDATE BINDING =='
for f in \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py \
.github/workflows/email-writing-review-service-tdd.yml
do
echo "--- $f"
cat -n "$f"
done
echo '== PR CHECK STATE =='
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 6319 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== SERVICE: CONSTANTS, ROLLBACK, CONTEXT, CANDIDATE VALIDATION =='
sed -n '1,347p' backend/services/email_writing_review_service.py | cat -n
echo '== SERVICE: PRIMARY REVIEW CONTROL FLOW =='
sed -n '348,727p' backend/services/email_writing_review_service.py | cat -n
echo '== SERVICE: FINALIZATION AND PERSISTENCE CONTROL FLOW =='
sed -n '728,1040p' backend/services/email_writing_review_service.py | cat -n
echo '== CANDIDATE VALIDATION IMPLEMENTATION =='
rg -n -A150 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_candidate_review.py \
backend/services/email_writing_review_service.py
echo '== JUDGE EXECUTION CONTRACT AND POLICY ADMISSION =='
rg -n -C10 \
'class .*Judge|def evaluate|def run_judge|asyncio\.to_thread|evaluate_policy_admission|evaluation_only|withhold|admit' \
backend/services/email_writing_judge.py \
backend/services/email_writing_policy.py \
backend/policies/email_writing_judge_evaluation_only_v1.json \
backend/services/email_writing_review_service.py
echo '== EVIDENCE MODEL =='
cat -n backend/db/email_writing_evidence.py
echo '== RUNTIME AUTHORITY SEARCH =='
rg -n -C5 -i \
'EmailWritingReviewService|email_writing_review_service|send_email|send.*email|publish.*diagnostic|review.*api' \
backend --glob '*.py'
echo '== FOCUSED TESTS: NAMES AND RELEVANT ASSERTIONS =='
for f in \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py
do
echo "--- $f"
rg -n -A80 -B8 \
'^async def test_|^def test_|block_commit|block_rollback|rollback_started|commit_started|CancelledError|review_evidence_unavailable|run_judge|selector|evidence_id|policy|send|publish' \
"$f"
done
echo '== TASK-9 FOCUSED WORKFLOW =='
cat -n .github/workflows/email-writing-review-service-tdd.yml
echo '== PR CHECKS (STATE ONLY) =='
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 146683 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== ROLLBACK HELPER AND BINDING VALIDATORS =='
sed -n '250,320p' backend/services/email_writing_review_service.py | cat -n
echo '== REVIEW ENTRY AND TIMEOUT/CANCELLATION HANDLERS =='
sed -n '340,435p' backend/services/email_writing_review_service.py | cat -n
echo '== AUTHORIZED-BUNDLE CANDIDATE/JUDGE/POLICY FLOW =='
sed -n '436,727p' backend/services/email_writing_review_service.py | cat -n
echo '== RESPONSE FINALIZATION AND PERSISTENCE =='
sed -n '728,1000p' backend/services/email_writing_review_service.py | cat -n
echo '== CANDIDATE SELECTOR/EVIDENCE VALIDATOR =='
ast-grep run --pattern 'def _validate_candidate_selectors_and_evidence($$$):
$$$' --lang python backend/services/email_writing_candidate_review.py
echo '== POLICY ADMISSION DEFINITION =='
ast-grep run --pattern 'def evaluate_policy_admission($$$):
$$$' --lang python backend/services/email_writing_policy.py
rg -n -C5 '"publish_decision"|"status"|evaluation_only|withhold' \
backend/policies/email_writing_judge_evaluation_only_v1.json
echo '== SERVICE-SPECIFIC RUNTIME REFERENCES =='
rg -n -C4 \
'from services\.email_writing_review_service import|import services\.email_writing_review_service|EmailWritingReviewService\s*\(' \
backend --glob '*.py' \
-g '!backend/tests/test_email_writing_review_service.py' \
-g '!backend/tests/test_email_writing_review_service_execution.py' \
-g '!backend/tests/test_email_writing_review_service_review_regressions.py' || true
echo '== EVIDENCE RECORD CONSTRUCTION FIELDS =='
sed -n '865,970p' backend/services/email_writing_review_service.py | cat -nLength of output: 35914
I found one issue.
The remaining requested boundaries are present in the reviewed source:
The checkout matches the requested commit. The You are interacting with an AI system. |
|
@coderabbitai review Please review exact current Task-9 head |
|
Tip For best results, initiate chat on the files or code changes. 🧩 Analysis chain🏁 Script executed: #!/bin/bash
set -euo pipefail
head='adb730e68ec7cdfc93c7f0a8fc151382e663c0d8'
base='21fa293d967bac0dd2b7b9a917b9fbce415134a2'
previous='88cffb54a981a3d7699e524947f81d7659bd71ad'
echo '== EXACT COMMIT, CHECKOUT, AND ANCESTRY =='
printf 'checkout='; git rev-parse HEAD
git show -s --format='commit=%H%nparents=%P%nsubject=%s' "$head"
git merge-base --is-ancestor "$base" "$head" && echo 'task8_predecessor_is_ancestor=yes' || echo 'task8_predecessor_is_ancestor=no'
git diff --quiet "$head" -- && echo 'checkout_matches_requested_head=yes' || echo 'checkout_matches_requested_head=no'
echo '== DELTA FROM IMMEDIATE PREDECESSOR =='
git diff --name-status "$previous" "$head"
echo '== SERVICE OUTLINE =='
ast-grep outline backend/services/email_writing_review_service.py --items all --type class,function --view expanded
echo '== SERVICE: ENTRY, DEADLINES, CANCELLATION, AND PERSISTENCE =='
nl -ba backend/services/email_writing_review_service.py | sed -n '1,470p'
nl -ba backend/services/email_writing_review_service.py | sed -n '700,1060p'
echo '== SERVICE: CANDIDATE, JUDGE, POLICY, AND RESPONSE FLOW =='
rg -n -C12 \
'_validate_candidate_result_binding|_validate_candidate_selectors_and_evidence|candidate_reviewer|run_judge|asyncio\.to_thread|evaluate_policy_admission|_persist_terminal_review|_persist_review|asyncio\.wait_for|asyncio\.timeout|CancelledError|TimeoutError|rollback|review_evidence_unavailable|diagnostic' \
backend/services/email_writing_review_service.py
echo '== CURRENT CANCELLATION REPAIR DIFF =='
git diff --find-renames --unified=120 "$previous" "$head" -- \
backend/services/email_writing_review_service.py \
backend/tests/test_email_writing_review_service_review_regressions.py \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
.github/workflows/email-writing-review-service-tdd.yml
echo '== CANDIDATE VALIDATOR AND JUDGE PORT =='
rg -n -A150 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_candidate_review.py \
backend/services/email_writing_review_service.py
rg -n -C10 \
'def run_judge|asyncio\.to_thread|class .*Judge|def evaluate' \
backend/services/email_writing_judge.py backend/services/email_writing_review_service.py
echo '== POLICY AND EVIDENCE MODEL =='
rg -n -C8 \
'def evaluate_policy_admission|evaluation_only|withhold|admit|publish_decision' \
backend/services/email_writing_policy.py backend/policies/email_writing_judge_evaluation_only_v1.json
cat -n backend/db/email_writing_evidence.py
echo '== FOCUSED REGRESSIONS =='
for f in \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py \
backend/tests/test_email_writing_review_service_review_regressions.py
do
echo "--- $f"
rg -n -A100 -B10 \
'^async def test_|^def test_|external.*cancel|CancelledError|cancel\(|commit_started|rollback_started|block_commit|block_rollback|review_evidence_unavailable|run_judge|selector|evidence_id|semantic output|diagnostic' \
"$f"
done
echo '== RUNTIME AUTHORITY REFERENCES =='
rg -n -C5 \
'from services\.email_writing_review_service import|import services\.email_writing_review_service|EmailWritingReviewService\s*\(|send_email|send.*email|publish.*diagnostic|review.*api' \
backend --glob '*.py' \
-g '!backend/tests/test_email_writing_review_service.py' \
-g '!backend/tests/test_email_writing_review_service_execution.py' \
-g '!backend/tests/test_email_writing_review_service_review_regressions.py' || true
echo '== TASK-9 WORKFLOW AND PR CHECKS =='
cat -n .github/workflows/email-writing-review-service-tdd.yml
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 7446 🏁 Script executed: #!/bin/bash
set -euo pipefail
head='adb730e68ec7cdfc93c7f0a8fc151382e663c0d8'
previous='88cffb54a981a3d7699e524947f81d7659bd71ad'
echo '== SERVICE: CONSTANTS, PORTS, VALIDATION, ROLLBACK =='
sed -n '1,347p' backend/services/email_writing_review_service.py | cat -n
echo '== SERVICE: REVIEW ENTRY, TIMEOUT, AND EXTERNAL CANCELLATION =='
sed -n '348,443p' backend/services/email_writing_review_service.py | cat -n
echo '== SERVICE: AUTHORIZED BUNDLE, CANDIDATE, JUDGE, AND POLICY =='
sed -n '444,735p' backend/services/email_writing_review_service.py | cat -n
echo '== SERVICE: FINALIZATION AND PERSISTENCE =='
sed -n '736,1040p' backend/services/email_writing_review_service.py | cat -n
echo '== CANCELLATION REPAIR DIFF =='
git diff --unified=120 "$previous" "$head" -- \
backend/services/email_writing_review_service.py \
backend/tests/test_email_writing_review_service_review_regressions.py
echo '== CANDIDATE VALIDATION IMPLEMENTATION =='
rg -n -A150 -B15 \
'^def _validate_candidate_selectors_and_evidence\b|^def _validate_candidate_result_binding\b' \
backend/services/email_writing_candidate_review.py \
backend/services/email_writing_review_service.py
echo '== JUDGE EXECUTOR AND POLICY ADMISSION =='
rg -n -C10 \
'def run_judge|asyncio\.to_thread|class .*Judge|def evaluate|def evaluate_policy_admission|evaluation_only|withhold|admit|publish_decision' \
backend/services/email_writing_judge.py \
backend/services/email_writing_review_service.py \
backend/services/email_writing_policy.py \
backend/policies/email_writing_judge_evaluation_only_v1.json
echo '== EVIDENCE MODEL =='
cat -n backend/db/email_writing_evidence.py
echo '== FOCUSED EXTERNAL-CANCELLATION AND PERSISTENCE TESTS =='
cat -n backend/tests/test_email_writing_review_service_review_regressions.py
for f in \
backend/tests/test_email_writing_review_service.py \
backend/tests/test_email_writing_review_service_execution.py
do
echo "--- $f"
rg -n -A90 -B10 \
'CancelledError|cancel\(|block_commit|block_rollback|commit_started|rollback_started|review_evidence_unavailable|run_judge|selector|evidence_id' \
"$f"
done
echo '== RUNTIME AUTHORITY REFERENCES =='
rg -n -C5 \
'from services\.email_writing_review_service import|import services\.email_writing_review_service|EmailWritingReviewService\s*\(|send_email|send.*email|publish.*diagnostic|review.*api' \
backend --glob '*.py' \
-g '!backend/tests/test_email_writing_review_service.py' \
-g '!backend/tests/test_email_writing_review_service_execution.py' \
-g '!backend/tests/test_email_writing_review_service_review_regressions.py' || true
echo '== WORKFLOW AND CURRENT CHECKS =='
cat -n .github/workflows/email-writing-review-service-tdd.yml
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 142056 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== REVIEW ENTRY AND EXTERNAL-CANCELLATION HANDLER =='
sed -n '340,445p' backend/services/email_writing_review_service.py | cat -n
echo '== AUTHORIZED CANDIDATE, JUDGE, AND POLICY FLOW =='
sed -n '444,705p' backend/services/email_writing_review_service.py | cat -n
echo '== TERMINAL FINALIZATION =='
sed -n '730,872p' backend/services/email_writing_review_service.py | cat -n
echo '== EVIDENCE PERSISTENCE AND BOUNDED ROLLBACK =='
sed -n '870,1045p' backend/services/email_writing_review_service.py | cat -n
echo '== VALIDATION HELPERS =='
sed -n '260,325p' backend/services/email_writing_review_service.py | cat -n
ast-grep run --pattern 'def _validate_candidate_selectors_and_evidence($$$):
$$$' --lang python backend/services/email_writing_candidate_review.py
echo '== CALL-SITE ORDER =='
rg -n -C8 \
'await self\.candidate_reviewer\.review|_validate_candidate_result_binding|await self\.judge_executor\.run_judge|self\.policy_evaluator|await self\._persist_review|await self\._persist_terminal_review' \
backend/services/email_writing_review_service.py
echo '== SERVICE-ONLY IMPORTS OR INSTANTIATIONS OUTSIDE TASK-9 TESTS =='
rg -n -C3 \
'from services\.email_writing_review_service import|import services\.email_writing_review_service|EmailWritingReviewService\s*\(' \
backend --glob '*.py' \
-g '!backend/tests/test_email_writing_review_service.py' \
-g '!backend/tests/test_email_writing_review_service_execution.py' \
-g '!backend/tests/test_email_writing_review_service_review_regressions.py' || true
echo '== TASK-9 EVIDENCE RECORD CONSTRUCTION =='
rg -n -A130 -B15 \
'EmailReviewSession\(|WritingDiagnosticRecord\(' \
backend/services/email_writing_review_service.py
echo '== TASK-9 LEXICAL-FALLBACK INDICATORS =='
rg -n -i -C3 \
'regex|keyword|phrase|sender.domain|recipient.count|language.name|nearest.text|fallback' \
backend/services/email_writing_review_service.py \
backend/services/email_writing_candidate_review.py \
backend/services/email_writing_judge.py || trueLength of output: 35312 🏁 Script executed: #!/bin/bash
set -u
echo '== CANDIDATE SELECTOR AND EVIDENCE VALIDATOR =='
rg -n -A150 -B10 \
'^def _validate_candidate_selectors_and_evidence\b' \
backend/services/email_writing_candidate_review.py || true
echo '== JUDGE EXECUTOR PORT AND SERVICE INVOCATION =='
sed -n '110,165p' backend/services/email_writing_review_service.py | cat -n
rg -n -C5 'asyncio\.to_thread|run_judge\(' \
backend/services/email_writing_review_service.py \
backend/services/email_writing_judge.py || true
echo '== POLICY ADMISSION IMPLEMENTATION AND CHECKED-IN POLICY =='
rg -n -A180 -B10 \
'^def evaluate_policy_admission\b' \
backend/services/email_writing_policy.py || true
rg -n -C5 \
'"status"|"publish_decision"|evaluation_only|withhold|admit' \
backend/policies/email_writing_judge_evaluation_only_v1.json || true
echo '== REVIEW-SERVICE RUNTIME INTEGRATION OUTSIDE TASK-9 TESTS =='
rg -n -C3 \
'email_writing_review_service|EmailWritingReviewService' \
backend --glob '*.py' \
-g '!backend/tests/test_email_writing_review_service.py' \
-g '!backend/tests/test_email_writing_review_service_execution.py' \
-g '!backend/tests/test_email_writing_review_service_review_regressions.py' || true
echo '== SERVICE-SPECIFIC LEXICAL SEMANTIC FALLBACK TERMS =='
rg -n -i -C3 \
'keyword|regex|phrase.list|sender.domain|recipient.count|language.name|nearest.text|semantic fallback' \
backend/services/email_writing_review_service.py || true
echo '== CURRENT CHECK STATE =='
gh pr checks 1535 --repo ContextualWisdomLab/naruon --json name,state,link,workflow 2>&1 || trueLength of output: 29051
I found no new actionable defects. The external-cancellation repair resolves the prior finding. In The new regression covers this path. It cancels an authorized review, blocks terminal evidence commit, confirms that commit receives cancellation, and confirms one rollback attempt. The existing timeout-state cleanup remains bounded. The Candidate trust boundary is correct:
The Judge separation remains correct. The service invokes the injected The policy path remains fail-closed. The evidence transaction stores owner scope, revision and projection identities, hashes, selectors, criterion-category receipts, policy/workflow identifiers, and admission state. It does not store raw draft text, replacements, explanations, prompts, or Judge output. I found no runtime import or instantiation of The checkout matches the requested commit. The Task-8 predecessor is an ancestor. The You are interacting with an AI system. |
|
PR governance metadata gate is not ready for
|
1 participant
Scope
Task 9 only: compose the already-separated Naruon email-writing boundaries into one bounded, fail-closed review service. This is a preparatory Draft lane stacked on the current Task-8 policy head. It does not add the HTTP review API, consume Inkspan, mutate/send email, publish a policy, or bypass the unresolved immutable fast-mlsirm artifact gate.
Runtime integration order remains Task 7 Judge → immutable fast-mlsirm pinning → Task 8 policy → Task 9 review service → Task 10 API. This slice may be developed and verified in parallel, but it must not merge as the canonical runtime descendant ahead of immutable dependency pinning.
Exact ancestry and TDD sequence
Fresh predecessor #1530 is
32c7e33c304d1b500ddc230044e2877edcf411c8. Exact current Task-9 head isc958b9964f925b6dca99e4c2c4db671934511a40. Current head was produced by an ordinary non-force two-parent restack after the Task-5 import-boundary repair propagated through Tasks 6–8; all predecessor checks/reviews are stale.Task-9-owned TDD sequence retained in the ancestry:
9af92412cf71fec51880637dd5fb14ab4274791c→ initial bounded review service38221826b004fc4f0646c645ca391ee42a00c4cd;9b5e015c02fc3e1c13403622ea2ea5e667083ffd→ bounded injected Judge execution fix9183d734ae98f47d60b96892489a29caf4c135f7;ac5ed3dd0fa15e4fb63e870b0e8f415083fd4987→ Candidate selector/evidence revalidation and bounded timeout-state evidence fixc421ab284fc07c1bc45f61f1a5a206d90cd40e33;214c1f5eb70ad5aa4ee970451fddca05f5aa347d→ bounded rollback cleanup88cffb54a981a3d7699e524947f81d7659bd71ad;5f3968bd64ac8e4e4748ec269bb77d674e01bc28→ cancellation evidence finalization fixadb730e68ec7cdfc93c7f0a8fc151382e663c0d8.No Jules/Cursor branch was modified, copied wholesale, force-updated, or used as writable ancestry. External #1402 remains read-only.
Review-service contract
The service executes only this order after server-authoritative context succeeds:
context build → Candidate reviewer → service-owned strict Candidate/bundle revalidation → bounded independent Judge execution → policy admission → deterministic revision/selector binding → one privacy-minimized evidence transaction → response.Key boundaries:
review_evidence_unavailable; no unrecorded diagnostic is returned;The checked-in Task-8 policy is evaluation-only/withhold, so this stack cannot expose user-facing diagnostics even if the pipeline executes successfully.
Immutable dependency boundary
Fresh inherited dependency truth remains: fast-mlsirm
v0.9.1contains the required Judge symbols in immutable tagged source but its GitHub release has no wheel/sdist asset. Naruon #1385 and fast-mlsirm #1692 are the current artifact/publication owner paths. No mutable fast-mlsirm or Inkspan branch, Git URL, source copy, local stub or workspace path is consumed here.Before runtime integration, the exact immutable fast-mlsirm distributable must be verified for version, integrity, source provenance, Python 3.14 installation/execution and Naruon hash lock; Task 7 must exercise those released symbols; then the policy/review-service lineage must be regenerated from that admitted predecessor. All predecessor checks/reviews are discarded when heads/bases move.
Current verification boundary
The dedicated
Email Writing Review Service TDDworkflow requires Python 3.14, the focused Task-9 tests, exact Task-9 statement+branch coverage at 100%, public docstrings, Ruff and compile checks on the current head.For exact current head
c958b9964f925b6dca99e4c2c4db671934511a40, pull-request job100237071157/ run33627024385is queued before executed steps. Same-head sibling job100237056302/ run33627020527terminated cancelled before useful execution. Both are non-passing; no predecessor run is transferable.Current review evidence must bind to
c958b9964f925b6dca99e4c2c4db671934511a40; prior reviews are predecessor evidence after the restack. Any new valid finding must be verified against current source and repaired test-first.Runner acquisition remains owned by
.github#712; the Naruon branch is not churned merely to retrigger that central condition.Keep Draft. Merge requires a future unchanged integration-current head to satisfy live rulesets/protection, every applicable CI/security/SAST/dependency/coverage/package/SBOM/provenance gate, zero valid unresolved threads, and any qualifying independent post-last-push approval actually required by live governance. No predecessor evidence, self-approval, bypass, force-push, destructive rebase, or gate weakening.