open-swe/tests/test_reviewer_findings.py

335 lines
12 KiB
Python
Raw Normal View History

feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
"""Unit tests for the Finding schema + thread-metadata helpers."""
from __future__ import annotations
from typing import Any
from unittest.mock import AsyncMock, patch
import pytest
from agent.reviewer_findings import (
SEVERITY_ORDER,
Finding,
append_finding,
filter_findings_for_publish,
list_findings,
mutate_findings,
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
new_finding,
new_finding_id,
replace_findings,
fix: reviewer publishes against stale head_sha on mid-run re-review (#1393) * fix: resolve reviewer head_sha from thread metadata, not frozen run config A push that lands while a reviewer run is in flight is delivered as a queued message into that run. The run's configurable is frozen at creation, so its head_sha still names the commit the run was created for — not the commit just pushed. publish_review then anchored the GitHub review to the stale commit and regressed last_reviewed_sha to it, and add_finding/update_finding stamped findings with the stale SHA. Persist the current head in thread metadata at every reviewer dispatch (both the ready-for-review and push paths, before they branch to create a run or queue a message), and add resolve_review_head_sha() which prefers the metadata head over the run config. Wire it into publish_review (review commit_id + last_reviewed_sha), add_finding (first_seen_sha) and update_finding (last_confirmed_sha). Falls back to the run config when metadata carries no head (first review, eval, tests). * fix: persist head_sha in manual review dispatch (trigger_pr_review_from_ref) resolve_review_head_sha prefers metadata[head_sha] over the run config, and the push/ready dispatchers write it — but trigger_pr_review_from_ref (Slack/GitHub @open-swe review, request_pr_review tool) created a run with a freshly-fetched config head while leaving metadata's head stale from a prior dispatch. A manual re-review at a newer commit would then resolve to the old head and publish/advance findings against it. Persist head_sha in that dispatch's metadata write too, so every run-creating reviewer dispatch keeps metadata in sync with the head its run targets. Caught by the Open SWE reviewer on this PR.
2026-06-03 11:38:56 -07:00
resolve_review_head_sha,
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
set_reviewer_thread_metadata,
update_finding_fields,
)
def _f(**overrides: Any) -> Finding:
base = new_finding(
severity="high",
feat: tune reviewer for precision — web/wiki tools + recalibrated prompt (#1312) * feat: tune reviewer for precision — web/wiki tools + recalibrated prompt Reviewer agent now has web_search, fetch_url, and http_request alongside the finding tools, so it can verify library semantics and consult the DeepWiki auto-generated wiki for public repos (https://deepwiki.com/<owner>/<repo>) before flagging cross-file or architectural concerns. Prompt rewritten to push precision over recall: - explicit severity ladder pushing reviews toward bimodal high/low instead of defaulting to medium - ≤200-char description target (gold set averages ~186 chars; we were at ~436) - mandatory docs / wiki / code lookup before flagging concurrency, security, or perf — the three categories that dominated false positives - "do not flag" list covering compiler/linter-catchable nits, speculative claims without a concrete attacker/interleaving/scale, style preferences the codebase doesn't share, and test-quality nits on non-test diffs - smart file-selection guidance for large PRs (deprioritize generated / vendored / pure-rename hunks) Eval config switched to openai:gpt-5.5 + high reasoning effort for the next benchmark run. * trim prompt * subagent prompting * confidence ratings * added medium * enforce confidence threshold * . * reviewer: precision-tuned prompt + drop confidence gate Rewrites the reviewer system prompt around a defensibility bar (anchor + failure mode + maintainer wouldn't say "not a bug"), an explicit do-not-file list (style nits, speculation, scope-policing, same-bug fan-out), and a checklist of 10 bug archetypes drawn from a per-PR audit of the eval golden set. The audit showed 145 FPs in the last eval split ~28% speculative, ~26% style-nit, ~31% real-but-unscored (mostly same-archetype fan-out); the new prompt targets each class directly. Confidence is still recorded on every finding for post-hoc calibration but no longer gates publication — the audit showed the gate was a no-op (agent self-rated 65% of findings "high" regardless), and the prompt's defensibility bar is the actual discipline. Drops CONFIDENCE_ORDER, CONFIDENCE_THRESHOLD, the confidence_threshold kwarg on filter_findings_for_publish, the confidence_filtered score_mode, and the min_confidence kwarg on the eval target's _extract_comments — all dead once the gate is gone. Also removes the "informational" severity tier from the Severity enum, SEVERITY_ORDER, and all validators / tests / docstrings. It was reserved for FYI observations the dataset never rewards. * benchmax * adding google provider * slight steering * tuning * more tuning * fix * cleanup * reducing overfitting * Add per-repo review style profiles and inject them into the reviewer. Dashboard users can analyze historical PR review feedback per repository, edit the resulting style guide, and have it loaded from LangGraph Store at reviewer runtime (including Martian eval runs) keyed by owner/name. Co-authored-by: Cursor <cursoragent@cursor.com> * Fix review style job errors leaking exception details to clients. Return generic dashboard messages while logging full stack traces server-side. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> Co-authored-by: Cursor <cursoragent@cursor.com>
2026-05-20 11:35:00 -07:00
confidence="high",
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
category="correctness",
file="foo.py",
start_line=10,
end_line=10,
description="boom",
sha="abc123",
)
base.update(overrides) # type: ignore[arg-type]
return base
def test_new_finding_id_format() -> None:
fid = new_finding_id()
assert fid.startswith("f_")
assert len(fid) == len("f_") + 10
def test_new_finding_defaults() -> None:
finding = _f()
assert finding["status"] == "open"
assert finding["side"] == "RIGHT"
assert finding["first_seen_sha"] == "abc123"
assert finding["last_confirmed_sha"] == "abc123"
assert finding["github_review_id"] is None
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
assert finding["github_review_comment_id"] is None
assert finding["github_review_comment_ids"] == []
assert finding["github_review_thread_id"] is None
assert finding["github_review_thread_ids"] == []
assert finding["github_review_run_id"] is None
assert finding["github_thread_resolved"] is False
assert finding["github_resolved_thread_ids"] == []
assert finding["last_human_reply_at"] is None
assert finding["resolution_note"] is None
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
assert finding["suggestion"] is None
def test_severity_order_monotonic() -> None:
assert (
feat: tune reviewer for precision — web/wiki tools + recalibrated prompt (#1312) * feat: tune reviewer for precision — web/wiki tools + recalibrated prompt Reviewer agent now has web_search, fetch_url, and http_request alongside the finding tools, so it can verify library semantics and consult the DeepWiki auto-generated wiki for public repos (https://deepwiki.com/<owner>/<repo>) before flagging cross-file or architectural concerns. Prompt rewritten to push precision over recall: - explicit severity ladder pushing reviews toward bimodal high/low instead of defaulting to medium - ≤200-char description target (gold set averages ~186 chars; we were at ~436) - mandatory docs / wiki / code lookup before flagging concurrency, security, or perf — the three categories that dominated false positives - "do not flag" list covering compiler/linter-catchable nits, speculative claims without a concrete attacker/interleaving/scale, style preferences the codebase doesn't share, and test-quality nits on non-test diffs - smart file-selection guidance for large PRs (deprioritize generated / vendored / pure-rename hunks) Eval config switched to openai:gpt-5.5 + high reasoning effort for the next benchmark run. * trim prompt * subagent prompting * confidence ratings * added medium * enforce confidence threshold * . * reviewer: precision-tuned prompt + drop confidence gate Rewrites the reviewer system prompt around a defensibility bar (anchor + failure mode + maintainer wouldn't say "not a bug"), an explicit do-not-file list (style nits, speculation, scope-policing, same-bug fan-out), and a checklist of 10 bug archetypes drawn from a per-PR audit of the eval golden set. The audit showed 145 FPs in the last eval split ~28% speculative, ~26% style-nit, ~31% real-but-unscored (mostly same-archetype fan-out); the new prompt targets each class directly. Confidence is still recorded on every finding for post-hoc calibration but no longer gates publication — the audit showed the gate was a no-op (agent self-rated 65% of findings "high" regardless), and the prompt's defensibility bar is the actual discipline. Drops CONFIDENCE_ORDER, CONFIDENCE_THRESHOLD, the confidence_threshold kwarg on filter_findings_for_publish, the confidence_filtered score_mode, and the min_confidence kwarg on the eval target's _extract_comments — all dead once the gate is gone. Also removes the "informational" severity tier from the Severity enum, SEVERITY_ORDER, and all validators / tests / docstrings. It was reserved for FYI observations the dataset never rewards. * benchmax * adding google provider * slight steering * tuning * more tuning * fix * cleanup * reducing overfitting * Add per-repo review style profiles and inject them into the reviewer. Dashboard users can analyze historical PR review feedback per repository, edit the resulting style guide, and have it loaded from LangGraph Store at reviewer runtime (including Martian eval runs) keyed by owner/name. Co-authored-by: Cursor <cursoragent@cursor.com> * Fix review style job errors leaking exception details to clients. Return generic dashboard messages while logging full stack traces server-side. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com> Co-authored-by: Cursor <cursoragent@cursor.com>
2026-05-20 11:35:00 -07:00
SEVERITY_ORDER["low"]
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
< SEVERITY_ORDER["medium"]
< SEVERITY_ORDER["high"]
< SEVERITY_ORDER["critical"]
)
def test_filter_findings_for_publish_drops_below_threshold_and_resolved() -> None:
findings = [
_f(id="f_a", severity="high", file="a.py", start_line=1, end_line=1),
_f(id="f_b", severity="low", file="b.py"),
_f(id="f_c", severity="critical", file="c.py", start_line=2, end_line=2),
_f(id="f_d", severity="high", file="d.py", status="resolved"),
]
surfaced = filter_findings_for_publish(findings, severity_threshold="medium", cap=10)
assert [f["id"] for f in surfaced] == ["f_c", "f_a"]
def test_filter_findings_for_publish_caps_results() -> None:
findings = [_f(id=f"f_{i}", severity="high", file=f"f{i}.py") for i in range(20)]
surfaced = filter_findings_for_publish(findings, severity_threshold="medium", cap=5)
assert len(surfaced) == 5
@pytest.mark.asyncio
async def test_list_findings_returns_empty_on_missing_metadata() -> None:
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
findings = await list_findings("tid")
assert findings == []
@pytest.mark.asyncio
async def test_list_findings_coerces_bad_entries() -> None:
fake_client = AsyncMock()
fake_client.threads.get.return_value = {
"metadata": {
"findings": [
{"id": "f_ok", "severity": "high", "file": "x.py"},
{"missing_id": True},
"not-a-dict",
]
}
}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
findings = await list_findings("tid")
assert [f["id"] for f in findings] == ["f_ok"]
@pytest.mark.asyncio
async def test_replace_findings_calls_threads_update() -> None:
fake_client = AsyncMock()
findings = [_f(id="f_x")]
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
await replace_findings("tid", findings)
fake_client.threads.update.assert_awaited_once_with(
thread_id="tid", metadata={"findings": findings}
)
@pytest.mark.asyncio
async def test_append_finding_appends_to_existing_list() -> None:
existing = _f(id="f_a")
new = _f(id="f_b")
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {"findings": [existing]}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
result = await append_finding("tid", new)
assert result["id"] == "f_b"
args = fake_client.threads.update.await_args
persisted = args.kwargs["metadata"]["findings"]
assert [f["id"] for f in persisted] == ["f_a", "f_b"]
@pytest.mark.asyncio
async def test_mutate_findings_reads_latest_before_mutating() -> None:
"""mutate_findings must operate on the freshest persisted list, not a stale
snapshot — the mutator receives whatever ``list_findings`` returns now."""
latest = [_f(id="f_fresh")]
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {"findings": latest}}
seen: list[str] = []
def _mutator(findings: list[Finding]) -> bool:
seen.extend(f["id"] for f in findings)
findings[0]["status"] = "resolved"
return True
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
result = await mutate_findings("tid", _mutator)
assert seen == ["f_fresh"]
assert result[0]["status"] == "resolved"
fake_client.threads.update.assert_awaited_once()
@pytest.mark.asyncio
async def test_mutate_findings_skips_write_when_unchanged() -> None:
"""A no-op mutation must NOT persist, so it can never clobber a concurrent
update that landed between the read and a would-be write."""
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {"findings": [_f(id="f_a")]}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
await mutate_findings("tid", lambda _findings: False)
fake_client.threads.update.assert_not_called()
feat: implement reviewer findings, publish_review, and watch mode (#1253) * feat: implement reviewer findings, publish_review, and watch mode Build out the reviewer agent end-to-end against the design in REVIEWER_DESIGN.md: - Findings as first-class state on the reviewer thread metadata (`agent/reviewer_findings.py`): Finding TypedDict with start_line/end_line ranges, suggestion text for ```suggestion blocks, github_review_comment_id for cross-run reconciliation, diff_hunk for UI rendering. Thread-level metadata gets `kind=reviewer`, `pr`, `last_reviewed_sha`, `watch` so a future frontend can list reviewer threads via the langgraph SDK. - Diff utilities (`agent/reviewer_diff.py`): parse_unified_diff, compute_diff_line_set for in-diff validation, extract_diff_hunk for caching the hunk on a Finding, compute_diff_in_sandbox for SHA-to-SHA diffs against the prepped repo. - Tools: `add_finding` (validates against the diff line set so out-of-diff ranges fail at creation, not at GitHub-publish), `update_finding`, `list_findings`, `publish_review`. The reviewer agent's tool list is swapped from `[]` (direct shell `gh api` calls) to these four. - Publish path (`agent/reviewer_publish.py` + `agent/tools/publish_review.py`): one POST /reviews call with body + inline comments + ```suggestion blocks, per-comment IDs stored back on findings, GraphQL `resolveReviewThread` fired for findings transitioning open->resolved on a re-review. - Reviewer graph: deterministic clone-or-fetch + checkout in the factory before the agent's first model call (warm- and cold-path symmetric); computed diff and in-diff line set passed via runnable config; system prompt rewritten for the single-evolving-findings model, severity ladder, in-diff-only discipline, and watch-mode reconciliation flow. - Watch mode in webapp.py: `push` event + `pull_request` closed/reopened added to supported events. New `process_github_push_event` resolves the open PR for the pushed branch, gates on the reviewer thread's `watch` flag, builds a re-review configurable, and triggers a run on the same canonical thread. `process_github_pr_close` toggles watch on closed/reopened. `set_reviewer_thread_metadata` is called on first review to install `kind=reviewer` + PR identity + watch=True. - Eval harness: target.py now extracts `add_finding` calls (mapped to the legacy {file, line, body, severity} shape the judge expects) and passes the right configurable so the prep step has base/head SHAs. - Tests: new unit suites for findings helpers, diff parsing, finding tools, publish rendering + GraphQL resolve, and watch-mode webhook handlers (push triggers re-review only when watching, idempotent on unchanged head SHA, PR close disables watch). Updated existing reviewer-webhook tests to mock `set_reviewer_thread_metadata`. - REVIEWER_EVAL_PLAN.md removed per user request; folded relevant context into REVIEWER_DESIGN.md. * fix(reviewer): correct git diff flags, scope, dedup, and review-comments URL Address PR #1253 review findings: - compute_diff_in_sandbox dropped the invalid `--no-prefix=false` flag (`option no-prefix takes no value` — every prep run was failing silently and the agent saw an empty diff). - compute_diff_in_sandbox grew a `merge_base` flag. First-review path now uses three-dot `base...head` (the merge-base diff GitHub renders on Files-changed) so we don't pick up changes that landed on the base branch after the PR diverged. Re-review delta keeps two-dot `last_reviewed_sha..head` since that's exactly the new commits. - publish_review skips findings that already carry `github_review_comment_id`. Without this, watched re-reviews re-posted every previously surfaced finding, and only the most-recent duplicate's id would later resolve when the issue got addressed. - fetch_review_comments URL now includes `{pull_number}` — `/repos/{owner}/{repo}/pulls/{pr_number}/reviews/{review_id}/comments` is the canonical endpoint; the old form 404s, so comment ids were never stored and watch-mode resolution couldn't run. Three new tests cover: three-dot vs two-dot wiring, no `--no-prefix` flag in the executed command, and that publish_review does not re-post findings whose `github_review_comment_id` is set. * fix(reviewer): default publish cap from 15 to 4 A clean PR with one critical issue padded out by three lower-severity findings is fine; fifteen is review spam. The agent can override per call when a PR genuinely warrants more.
2026-05-07 14:48:43 -07:00
@pytest.mark.asyncio
async def test_update_finding_fields_mutates_only_target() -> None:
a = _f(id="f_a", description="orig-a")
b = _f(id="f_b", description="orig-b")
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {"findings": [a, b]}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
updated = await update_finding_fields("tid", "f_b", {"status": "resolved"})
assert updated is not None
assert updated["status"] == "resolved"
persisted = fake_client.threads.update.await_args.kwargs["metadata"]["findings"]
by_id = {f["id"]: f for f in persisted}
assert by_id["f_a"]["status"] == "open"
assert by_id["f_b"]["status"] == "resolved"
@pytest.mark.asyncio
async def test_update_finding_fields_returns_none_for_unknown_id() -> None:
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {"findings": [_f(id="f_a")]}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
result = await update_finding_fields("tid", "f_missing", {"status": "resolved"})
assert result is None
fake_client.threads.update.assert_not_called()
@pytest.mark.asyncio
async def test_set_reviewer_thread_metadata_includes_kind() -> None:
fake_client = AsyncMock()
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
await set_reviewer_thread_metadata("tid", watch=True, last_reviewed_sha="sha")
metadata = fake_client.threads.update.await_args.kwargs["metadata"]
assert metadata["kind"] == "reviewer"
assert metadata["watch"] is True
assert metadata["last_reviewed_sha"] == "sha"
assert "pr" not in metadata
assert "findings" not in metadata
fix: reviewer publishes against stale head_sha on mid-run re-review (#1393) * fix: resolve reviewer head_sha from thread metadata, not frozen run config A push that lands while a reviewer run is in flight is delivered as a queued message into that run. The run's configurable is frozen at creation, so its head_sha still names the commit the run was created for — not the commit just pushed. publish_review then anchored the GitHub review to the stale commit and regressed last_reviewed_sha to it, and add_finding/update_finding stamped findings with the stale SHA. Persist the current head in thread metadata at every reviewer dispatch (both the ready-for-review and push paths, before they branch to create a run or queue a message), and add resolve_review_head_sha() which prefers the metadata head over the run config. Wire it into publish_review (review commit_id + last_reviewed_sha), add_finding (first_seen_sha) and update_finding (last_confirmed_sha). Falls back to the run config when metadata carries no head (first review, eval, tests). * fix: persist head_sha in manual review dispatch (trigger_pr_review_from_ref) resolve_review_head_sha prefers metadata[head_sha] over the run config, and the push/ready dispatchers write it — but trigger_pr_review_from_ref (Slack/GitHub @open-swe review, request_pr_review tool) created a run with a freshly-fetched config head while leaving metadata's head stale from a prior dispatch. A manual re-review at a newer commit would then resolve to the old head and publish/advance findings against it. Persist head_sha in that dispatch's metadata write too, so every run-creating reviewer dispatch keeps metadata in sync with the head its run targets. Caught by the Open SWE reviewer on this PR.
2026-06-03 11:38:56 -07:00
@pytest.mark.asyncio
async def test_set_reviewer_thread_metadata_persists_head_sha() -> None:
fake_client = AsyncMock()
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
await set_reviewer_thread_metadata("tid", head_sha="newhead")
metadata = fake_client.threads.update.await_args.kwargs["metadata"]
assert metadata["head_sha"] == "newhead"
@pytest.mark.asyncio
async def test_resolve_review_head_sha_prefers_metadata_over_config() -> None:
"""A mid-run push records the live head in thread metadata; it must win over
the stale head frozen in the run's config."""
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {"head_sha": "metahead"}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
head = await resolve_review_head_sha("tid", {"head_sha": "confighead"})
assert head == "metahead"
@pytest.mark.asyncio
async def test_resolve_review_head_sha_falls_back_to_config_when_metadata_empty() -> None:
fake_client = AsyncMock()
fake_client.threads.get.return_value = {"metadata": {}}
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
head = await resolve_review_head_sha("tid", {"head_sha": "confighead"})
assert head == "confighead"
@pytest.mark.asyncio
async def test_resolve_review_head_sha_falls_back_without_thread_id() -> None:
fake_client = AsyncMock()
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
head = await resolve_review_head_sha("", {"head_sha": "confighead"})
assert head == "confighead"
fake_client.threads.get.assert_not_called()
@pytest.mark.asyncio
async def test_replace_findings_raises_domain_error_when_thread_missing() -> None:
import httpx
from langgraph_sdk.errors import NotFoundError
from agent.reviewer_findings import ReviewerThreadMissingError
not_found = NotFoundError(
"thread tid not found",
response=httpx.Response(404, request=httpx.Request("PATCH", "http://x")),
body=None,
)
fake_client = AsyncMock()
fake_client.threads.update.side_effect = not_found
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
with pytest.raises(ReviewerThreadMissingError) as excinfo:
await replace_findings("tid", [_f(id="f_a")])
assert excinfo.value.thread_id == "tid"
assert "not found" in str(excinfo.value)
def _not_found(method: str = "GET") -> Exception:
import httpx
from langgraph_sdk.errors import NotFoundError
return NotFoundError(
"thread tid not found",
response=httpx.Response(404, request=httpx.Request(method, "http://x")),
body=None,
)
@pytest.mark.asyncio
async def test_get_thread_metadata_raises_domain_error_when_thread_missing() -> None:
"""A missing thread must surface as ReviewerThreadMissingError, not be
swallowed into ``{}`` — that produced misleading tool results like
"No finding found" instead of the do-not-retry contract."""
from agent.reviewer_findings import ReviewerThreadMissingError, get_thread_metadata
fake_client = AsyncMock()
fake_client.threads.get.side_effect = _not_found()
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
with pytest.raises(ReviewerThreadMissingError):
await get_thread_metadata("tid")
@pytest.mark.asyncio
async def test_get_thread_metadata_still_degrades_on_other_failures() -> None:
from agent.reviewer_findings import get_thread_metadata
fake_client = AsyncMock()
fake_client.threads.get.side_effect = RuntimeError("transient")
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
assert await get_thread_metadata("tid") == {}
@pytest.mark.asyncio
async def test_set_reviewer_thread_metadata_raises_domain_error_when_thread_missing() -> None:
from agent.reviewer_findings import (
ReviewerThreadMissingError,
set_reviewer_thread_metadata,
)
fake_client = AsyncMock()
fake_client.threads.update.side_effect = _not_found("PATCH")
with patch("agent.reviewer_findings.get_client", return_value=fake_client):
with pytest.raises(ReviewerThreadMissingError):
await set_reviewer_thread_metadata("tid", last_reviewed_sha="sha")