← Back to all submissions
{% if sub.status in ['pending_inline','pending_llm'] %}
Review in progress — checks running against the bundle.
Page auto-refreshes when verdict lands.
0s
{% endif %}
{# ── Version switcher ────────────────────────────────────────────────────── #}
{# All submissions linked to this entity, newest first. Admin clicks
any row to jump to its detail. Surfaces multi-version entities so
verdicts across versions can be compared without bouncing back to
the queue. #}
{% if sibling_submissions and sibling_submissions|length > 1 %}
Submissions for this entity ({{ sibling_submissions|length }})
| v# |
Status |
Hash |
Reviewed by |
Created |
|
{% for s in sibling_submissions %}
|
{% if s.version_no %}
v{{ s.version_no }}
{% else %}
—
{% endif %}
{% if s.version_no and s.version_no == entity_version_no %}
current
{% endif %}
|
{{ s.status }} |
{{ (s.version or '')[:12] }} |
{{ s.reviewed_by_model or '—' }} |
{{ s.created_at.strftime("%Y-%m-%d %H:%M") if s.created_at else "" }}
|
{% if s.is_current %}
viewing
{% else %}
Open →
{% endif %}
|
{% endfor %}
{% endif %}
{# ── Override history ────────────────────────────────────────────────────── #}
{% if sub.override_by %}
Override
Overridden by {{ override_email or sub.override_by }}{% if sub.updated_at %} on {{ sub.updated_at.strftime("%Y-%m-%d %H:%M:%S UTC") }}{% endif %}.
{% if sub.override_reason %}
Reason: {{ sub.override_reason }}{% endif %}
{% endif %}
{# ── Activity timeline ───────────────────────────────────────────────────── #}
{# v37 split: per-submission events surface here. Entity-wide events
(archive / install / hard-delete) — which apply to ALL versions of
this entity — surface in the separate "Entity activity" card below
so admin can tell version-scoped from entity-scoped at a glance. #}
{% macro _render_timeline(rows) %}
{% for row in rows %}
-
{{ row.timestamp.strftime("%Y-%m-%d %H:%M:%S") if row.timestamp else "" }}
{{ row.action }}
{# Inline `vN` chip when the audit's params reference a
specific version_no — makes it clear in the entity-wide
timeline which version each event applied to. #}
{% if row.params and row.params is mapping and row.params.version_no %}
v{{ row.params.version_no }}
{% endif %}
{% if row.actor_email %} · {{ row.actor_email }}{% endif %}
{% if row.params and row.params is mapping %}
{% if row.params.outcome %} · → {{ row.params.outcome }}{% endif %}
{% if row.params.risk_level %} · risk={{ row.params.risk_level }}{% endif %}
{% if row.params.reason %} · "{{ row.params.reason[:80] }}"{% endif %}
{% if row.params.renamed_to %} · → {{ row.params.renamed_to }}{% endif %}
{% if row.params.restored_from_version_no %} · from v{{ row.params.restored_from_version_no }}{% endif %}
{% endif %}
just now
{% endfor %}
{% endmacro %}
{% if submission_audit_rows %}
Activity timeline{% if submission_version_no %} v{{ submission_version_no }}{% endif %}
Recorded actions for THIS submission row only — accept / verdict /
rescan / override / retry / bundle download. Use this to confirm a
verdict is fresh (timestamp on the latest store.submission.rescan
row matches the time you clicked Rescan).
{{ _render_timeline(submission_audit_rows) }}
{% endif %}
{% if entity_audit_rows %}
Entity activity (all versions)
Entity-wide events that apply to every version of this entity —
creation, archive, install / uninstall, hard delete. Same content
shows on every submission for this entity by design.
{{ _render_timeline(entity_audit_rows) }}
{% endif %}
{# ── Inline checks ───────────────────────────────────────────────────────── #}
{# Each ``det-card`` below renders verdict data scoped to THIS
submission row (i.e. one specific version). Section headers get
a ``vN`` chip when ``submission_version_no`` is known so admin
never confuses one version's verdict for another's. #}
{% set ic = sub.inline_checks or {} %}
{% macro _vchip() %}{% if submission_version_no %}
v{{ submission_version_no }}{% endif %}{% endmacro %}
Manifest check{{ _vchip() }} {% if ic.manifest %}{{ ic.manifest.status or 'pass' }}{% endif %}
{% if ic.manifest and ic.manifest.issues %}
{% for issue in ic.manifest.issues %}{{ issue }}
{% else %}
No manifest issues.
{% endif %}
Static security scan{{ _vchip() }} {% if ic.static_security %}{{ ic.static_security.status or 'pass' }}{% endif %}
{% if ic.static_security and ic.static_security.findings %}
| Severity | Category | File:line | Reason | Snippet |
|---|
{% for f in ic.static_security.findings %}
| {{ f.severity }} |
{{ f.category }} |
{{ f.file }}:{{ f.line }} |
{{ f.reason }} |
{{ f.snippet }} |
{% endfor %}
{% else %}
No security findings.
{% endif %}
Quality + templating{{ _vchip() }} {% if ic.quality %}{{ ic.quality.status or 'pass' }}{% endif %}
{% if ic.quality %}
Template placeholders found: {{ ic.quality.template_placeholders or 0 }}
{% if ic.quality.issues %}
{% for issue in ic.quality.issues %}{{ issue }}
{% endif %}
{% if ic.quality.template_recommendation %}
{{ ic.quality.template_recommendation }}
{% endif %}
{% else %}
No quality data.
{% endif %}
{# ── LLM review ──────────────────────────────────────────────────────────── #}
LLM security review{{ _vchip() }}
{% if sub.llm_findings and sub.llm_findings.risk_level %}
{{ sub.llm_findings.risk_level }}
{% endif %}
{% if sub.llm_findings %}
{% if sub.llm_findings.error %}
Error: {{ sub.llm_findings.error }}
{% else %}
{% if sub.llm_findings.summary %}
{{ sub.llm_findings.summary }}
{% endif %}
{% if sub.llm_findings.findings %}
Security findings
| Severity | Category | File | Explanation | Fix hint |
|---|
{% for f in sub.llm_findings.findings %}
| {{ f.severity }} |
{{ f.category }} |
{{ f.file }} |
{{ f.explanation }} |
{{ f.fix_hint or "" }} |
{% endfor %}
{% endif %}
{# Content-quality issues. Pre-fix the template only rendered
the security findings table, so a submission blocked
purely on content_quality (e.g. weak description) showed
up as 'No findings — model verdict was clean' even though
status was blocked_llm. Surfaced live by an admin
investigating a blocked re-restore of an already-approved
bundle. #}
{% set cq = sub.llm_findings.content_quality %}
{% if cq and cq.issues %}
Content quality issues
{{ cq.verdict }}
| File | Field | Issue | Suggested rewrite |
|---|
{% for issue in cq.issues %}
{{ issue.file }} |
{{ issue.field }} |
{{ issue.issue }} |
{{ issue.hint or "" }} |
{% endfor %}
{% endif %}
{% if not sub.llm_findings.findings and not (cq and cq.issues) %}
{% if sub.status in ['blocked_llm', 'review_error'] %}
Blocked but no findings recorded.
This is usually a transient LLM non-determinism: the
reviewer returned a fail verdict without enumerating
the offending items. Click Rescan to re-run
the pipeline, or Override with a written
reason if you've verified the bundle offline.
{% else %}
No findings — model verdict was clean.
{% endif %}
{% endif %}
{% if sub.llm_findings.reused_from_submission_id %}
{% endif %}
{% endif %}
{% elif sub.status in ['pending_inline', 'pending_llm'] %}
Review still in progress…
{% else %}
No LLM verdict (review skipped or not yet run).
{% endif %}
{# ── Other attempts by submitter ─────────────────────────────────────────── #}
{% if other_count %}
{% endif %}
{% endblock %}