Bind C-25 JSON to the scanned source commit (STATE-WP-0094-T03)
All checks were successful
CI Smoke / host-smoke (push) Successful in 0s
CI Smoke / container-smoke (push) Successful in 1s
Build and Publish Multi-Context Image / build-and-push (push) Successful in 32s

The checker JSON reports per repository repo_slug, source_revision (full
40-hex HEAD SHA) and source_revision_reason; C-25 findings repeat them so a
findings-only consumer can bind. A dirty workplans/ tree, a non-git path,
a missing HEAD or a tree changing mid-scan yields null plus a reason, never
HEAD or a branch name. plan_owner_returns validates and passes the revision
into every candidate and refuses findings bound elsewhere.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 692292@bnt-lap001
Assistant-Session: 89c7d8e8-e1b3-47eb-9411-a0ab227fca2f
This commit is contained in:
tegwick 2026-10-04 13:33:24 +02:00
parent a0083ab78c
commit 150cb0b201
6 changed files with 284 additions and 1 deletions

View file

@ -312,6 +312,10 @@ class ConsistencyReport:
repo_path: str
issues: list[Issue] = field(default_factory=list)
fixes_applied: list[str] = field(default_factory=list)
# Commit the workplans/ scan read (STATE-WP-0094): a full 40-hex SHA, or
# None with a machine-readable reason. Never a branch name or a guess.
source_revision: str | None = None
source_revision_reason: str | None = "not_scanned"
def add(self, **kwargs) -> Issue:
issue = Issue(**kwargs)
@ -1465,6 +1469,9 @@ def check_repo(
_check_orphan_db(api_base, repo_id, set(), report)
return report
# Bind the scan to the commit it reads; re-verified after C-25 below.
report.source_revision, report.source_revision_reason = scan_source_revision(repo_dir)
# Parse workplan files
workplan_infos: list[tuple[Path, dict, str]] = []
file_ws_ids: dict[str, tuple[Path, dict, str]] = {} # ws_id → (file, meta, body)
@ -2429,6 +2436,10 @@ def check_repo(
_check_ghost_duplicates(api_base, workplan_infos, file_ws_ids, report)
_check_blocked_unblock_sweep(api_base, repo_slug, repo_dir, workplan_infos, report)
if report.source_revision is not None:
# The tree must not have moved while the files were read and C-25 ran.
if scan_source_revision(repo_dir) != (report.source_revision, None):
report.source_revision, report.source_revision_reason = None, "changed_during_scan"
_check_inbox_hygiene(api_base, repo_slug, report)
_check_workplan_prefixes(repo_dir, repo_slug, workplan_infos, report)
_check_workplan_id_collisions(api_base, repo_slug, repo_dir, workplan_infos, report)
@ -2786,6 +2797,44 @@ def collect_inbox_hygiene(api_base: str, repo_slug: str) -> dict[str, Any]:
}
_FULL_SHA = re.compile(r"^[0-9a-f]{40}$")
def scan_source_revision(repo_dir: Path) -> tuple[str | None, str | None]:
"""Return ``(sha, None)`` for a clean ``workplans/`` tree, else ``(None, reason)``.
``sha`` is the full 40-hex HEAD commit. Any change under ``workplans/``
(modified, staged, untracked or ignored) means the scanned files are not
that commit, so no revision is claimed. Reasons: ``not_a_git_tree``,
``no_head_commit``, ``unsupported_object_format``, ``dirty_workplans``,
``git_status_failed``.
"""
def git(*args: str) -> str | None:
try:
return subprocess.run(
["git", *args], cwd=repo_dir, capture_output=True, text=True,
timeout=30, check=True,
).stdout
except (subprocess.CalledProcessError, subprocess.TimeoutExpired, OSError):
return None
if not Path(repo_dir).is_dir() or git("rev-parse", "--is-inside-work-tree") is None:
return None, "not_a_git_tree"
head = git("rev-parse", "--verify", "--quiet", "HEAD^{commit}")
if head is None or not head.strip():
return None, "no_head_commit"
sha = head.strip()
if not _FULL_SHA.fullmatch(sha):
return None, "unsupported_object_format"
status = git("status", "--porcelain", "--untracked-files=all", "--ignored",
"--", "workplans")
if status is None:
return None, "git_status_failed"
if status.strip():
return None, "dirty_workplans"
return sha, None
C25_PAGE_SIZE = 200
C25_MAX_PAGES = 50
@ -4360,9 +4409,14 @@ def render_text(report: ConsistencyReport, show_info: bool = True) -> str:
def report_to_dict(report: ConsistencyReport) -> dict:
revision = {
"source_revision": report.source_revision,
"source_revision_reason": report.source_revision_reason,
}
return {
"repo_slug": report.repo_slug,
"repo_path": report.repo_path,
**revision,
"issues": [
{
"severity": i.severity,
@ -4373,6 +4427,7 @@ def report_to_dict(report: ConsistencyReport) -> dict:
"file_value": i.file_value,
"db_value": i.db_value,
"fixable": i.fixable,
**({"repo_slug": report.repo_slug, **revision} if i.check_id == "C-25" else {}),
**({"owner_return": i._fix_context} if i.check_id == "C-25" and i._fix_context.get("task_record_id") else {}),
}
for i in report.issues