Harden workflow snapshot rehydration

Author: LEANDERANTONY

backend/services/workspace_run_jobs.py

@@ -7,11 +7,13 @@
 from typing import Any
 
 from src.errors import AppError
+from src.logging_utils import get_logger, log_event
 
 from backend.services.workspace_service import run_workspace_analysis
 
 
 JOB_TTL_SECONDS = 60 * 30
+LOGGER = get_logger(__name__)
 
 
 @dataclass
@@ -102,6 +104,15 @@ def _run_job(
             job.updated_at = time.time()
     except AppError as error:
         message = error.user_message
+        log_event(
+            LOGGER,
+            30,
+            "workspace_run_job_failed",
+            "The background workspace analysis job failed with an application error.",
+            job_id=job_id,
+            error_type=type(error).__name__,
+            message=message,
+        )
         with _LOCK:
             job = _JOBS.get(job_id)
             if job is None:
@@ -110,6 +121,7 @@ def _run_job(
             job.error_message = message
             job.updated_at = time.time()
     except Exception as error:  # pragma: no cover - defensive server fallback
+        LOGGER.exception("Background workspace analysis job crashed.", extra={"job_id": job_id})
         with _LOCK:
             job = _JOBS.get(job_id)
             if job is None:

src/workflow_payloads.py

@@ -286,6 +286,10 @@ def build_saved_workflow_snapshot_from_data(payload: dict):
     )
 
 
+def _as_dict(payload):
+    return payload if isinstance(payload, dict) else {}
+
+
 def _build_imported_job_posting(payload):
     if not isinstance(payload, dict):
         return None
@@ -309,6 +313,7 @@ def _build_imported_job_posting(payload):
 
 
 def _build_candidate_profile(payload: dict):
+    payload = _as_dict(payload)
     return CandidateProfile(
         full_name=str(payload.get("full_name", "") or ""),
         location=str(payload.get("location", "") or ""),
@@ -324,6 +329,7 @@ def _build_candidate_profile(payload: dict):
 
 
 def _build_work_experience(payload: dict):
+    payload = _as_dict(payload)
     return WorkExperience(
         title=str(payload.get("title", "") or ""),
         organization=str(payload.get("organization", "") or ""),
@@ -335,6 +341,7 @@ def _build_work_experience(payload: dict):
 
 
 def _build_education_entry(payload: dict):
+    payload = _as_dict(payload)
     return EducationEntry(
         institution=str(payload.get("institution", "") or ""),
         degree=str(payload.get("degree", "") or ""),
@@ -345,7 +352,9 @@ def _build_education_entry(payload: dict):
 
 
 def _build_job_description(payload: dict):
+    payload = _as_dict(payload)
     requirements = payload.get("requirements") or {}
+    requirements = _as_dict(requirements)
     return JobDescription(
         title=str(payload.get("title", "") or ""),
         raw_text=str(payload.get("raw_text", "") or ""),
@@ -362,6 +371,7 @@ def _build_job_description(payload: dict):
 
 
 def _build_fit_analysis(payload: dict):
+    payload = _as_dict(payload)
     return FitAnalysis(
         target_role=str(payload.get("target_role", "") or ""),
         overall_score=int(payload.get("overall_score", 0) or 0),
@@ -378,6 +388,7 @@ def _build_fit_analysis(payload: dict):
 
 
 def _build_tailored_draft(payload: dict):
+    payload = _as_dict(payload)
     return TailoredResumeDraft(
         target_role=str(payload.get("target_role", "") or ""),
         professional_summary=str(payload.get("professional_summary", "") or ""),
@@ -388,6 +399,7 @@ def _build_tailored_draft(payload: dict):
 
 
 def _build_profile_output(payload: dict):
+    payload = _as_dict(payload)
     return ProfileAgentOutput(
         positioning_headline=str(payload.get("positioning_headline", "") or ""),
         evidence_highlights=[str(item) for item in payload.get("evidence_highlights", []) or []],
@@ -397,6 +409,7 @@ def _build_profile_output(payload: dict):
 
 
 def _build_job_output(payload: dict):
+    payload = _as_dict(payload)
     return JobAgentOutput(
         requirement_summary=str(payload.get("requirement_summary", "") or ""),
         priority_skills=[str(item) for item in payload.get("priority_skills", []) or []],
@@ -406,6 +419,7 @@ def _build_job_output(payload: dict):
 
 
 def _build_fit_output(payload: dict):
+    payload = _as_dict(payload)
     return FitAgentOutput(
         fit_summary=str(payload.get("fit_summary", "") or ""),
         top_matches=[str(item) for item in payload.get("top_matches", []) or []],
@@ -414,6 +428,7 @@ def _build_fit_output(payload: dict):
 
 
 def _build_tailoring_output(payload: dict):
+    payload = _as_dict(payload)
     return TailoringAgentOutput(
         professional_summary=str(payload.get("professional_summary", "") or ""),
         rewritten_bullets=[str(item) for item in payload.get("rewritten_bullets", []) or []],
@@ -425,6 +440,7 @@ def _build_tailoring_output(payload: dict):
 def _build_strategy_output(payload):
     if not payload:
         return None
+    payload = _as_dict(payload)
     return StrategyAgentOutput(
         recruiter_positioning=str(payload.get("recruiter_positioning", "") or ""),
         cover_letter_talking_points=[str(item) for item in payload.get("cover_letter_talking_points", []) or []],
@@ -433,6 +449,7 @@ def _build_strategy_output(payload):
 
 
 def _build_review_output(payload: dict):
+    payload = _as_dict(payload)
     return ReviewAgentOutput(
         approved=bool(payload.get("approved", False)),
         grounding_issues=[str(item) for item in payload.get("grounding_issues", []) or []],
@@ -447,6 +464,7 @@ def _build_review_output(payload: dict):
 def _build_resume_generation_output(payload):
     if not payload:
         return None
+    payload = _as_dict(payload)
     return ResumeGenerationAgentOutput(
         professional_summary=str(payload.get("professional_summary", "") or ""),
         highlighted_skills=[str(item) for item in payload.get("highlighted_skills", []) or []],
@@ -459,6 +477,7 @@ def _build_resume_generation_output(payload):
 def _build_cover_letter_output(payload):
     if not payload:
         return None
+    payload = _as_dict(payload)
     return CoverLetterAgentOutput(
         greeting=str(payload.get("greeting", "") or ""),
         opening_paragraph=str(payload.get("opening_paragraph", "") or ""),
@@ -470,6 +489,7 @@ def _build_cover_letter_output(payload):
 
 
 def _build_review_pass_result(payload: dict):
+    payload = _as_dict(payload)
     return ReviewPassResult(
         pass_index=int(payload.get("pass_index", 0) or 0),
         tailoring=_build_tailoring_output(payload.get("tailoring") or {}),
@@ -481,6 +501,7 @@ def _build_review_pass_result(payload: dict):
 def _build_agent_result(payload):
     if not payload:
         return None
+    payload = _as_dict(payload)
     return AgentWorkflowResult(
         mode=str(payload.get("mode", "") or ""),
         model=str(payload.get("model", "") or ""),

tests/test_workflow_payloads.py

@@ -0,0 +1,106 @@
+from backend.services.artifact_export_service import preview_workspace_artifact
+from src.workflow_payloads import build_saved_workflow_snapshot_from_data
+
+
+def _malformed_workspace_snapshot():
+    return {
+        "candidate_profile": {
+            "full_name": "Leander Antony",
+            "location": "Chennai",
+            "contact_lines": ["leander@example.com"],
+            "source": "workspace",
+            "resume_text": "Resume text",
+            "skills": ["Python", "FastAPI"],
+            "experience": [
+                {
+                    "title": "Engineer",
+                    "organization": "Example Corp",
+                    "location": "Remote",
+                    "description": "Built features",
+                    "start": "2023",
+                    "end": "2024",
+                },
+                "unexpected-string-entry",
+            ],
+            "education": [None],
+            "certifications": [],
+            "source_signals": [],
+        },
+        "job_description": {
+            "title": "ML Engineer",
+            "raw_text": "We need Python and FastAPI experience.",
+            "cleaned_text": "We need Python and FastAPI experience.",
+            "location": "Remote",
+            "requirements": {
+                "hard_skills": ["Python", "FastAPI"],
+                "soft_skills": ["Communication"],
+                "experience_requirement": "3+ years",
+                "must_haves": ["APIs"],
+                "nice_to_haves": ["LLMs"],
+            },
+        },
+        "fit_analysis": {
+            "target_role": "ML Engineer",
+            "overall_score": 82,
+            "readiness_label": "Strong",
+            "matched_hard_skills": ["Python"],
+            "missing_hard_skills": ["Kubernetes"],
+            "matched_soft_skills": ["Communication"],
+            "missing_soft_skills": [],
+            "experience_signal": "Relevant backend work",
+            "strengths": ["API design"],
+            "gaps": ["Kubernetes"],
+            "recommendations": ["Highlight deployment work"],
+        },
+        "tailored_draft": {
+            "target_role": "ML Engineer",
+            "professional_summary": "Backend engineer focused on ML tooling.",
+            "highlighted_skills": ["Python", "FastAPI"],
+            "priority_bullets": ["Built APIs for AI workflows"],
+            "gap_mitigation_steps": ["Call out transferable infra work"],
+        },
+        "agent_result": {
+            "mode": "agentic",
+            "model": "gpt-test",
+            "profile": {"positioning_headline": "ML engineer", "evidence_highlights": [], "strengths": [], "cautions": []},
+            "job": {"requirement_summary": "Python-heavy role", "priority_skills": [], "must_have_themes": [], "messaging_guidance": []},
+            "fit": {"fit_summary": "Good fit", "top_matches": [], "key_gaps": []},
+            "tailoring": {
+                "professional_summary": "Backend engineer focused on ML tooling.",
+                "rewritten_bullets": ["Built APIs for AI workflows"],
+                "highlighted_skills": ["Python", "FastAPI"],
+                "cover_letter_themes": ["ownership"],
+            },
+            "review": {
+                "approved": True,
+                "grounding_issues": [],
+                "unresolved_issues": [],
+                "revision_requests": [],
+                "final_notes": [],
+            },
+            "review_history": [None, "unexpected-review-entry"],
+        },
+        "imported_job_posting": None,
+    }
+
+
+def test_build_saved_workflow_snapshot_tolerates_malformed_nested_items():
+    snapshot = build_saved_workflow_snapshot_from_data(_malformed_workspace_snapshot())
+
+    assert snapshot is not None
+    assert snapshot.candidate_profile.experience[1].title == ""
+    assert snapshot.candidate_profile.education[0].institution == ""
+    assert snapshot.agent_result is not None
+    assert snapshot.agent_result.review_history[0].pass_index == 0
+    assert snapshot.agent_result.review_history[1].pass_index == 0
+
+
+def test_preview_workspace_artifact_tolerates_malformed_nested_items():
+    response = preview_workspace_artifact(
+        workspace_snapshot=_malformed_workspace_snapshot(),
+        artifact_kind="tailored_resume",
+        resume_theme="classic_ats",
+    )
+
+    assert response["status"] == "ready"
+    assert "<html" in response["html"].lower()