feat: add retry endpoint for jobs, update frontend to support retry functionality

backend/app/api/v1/jobs.py

@@ -8,7 +8,7 @@
 import asyncio
 from fastapi import APIRouter, Depends, HTTPException, status
 from fastapi.responses import StreamingResponse, Response
-from sqlalchemy import select
+from sqlalchemy import select, update
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.api.deps import get_db
@@ -128,6 +128,66 @@ async def delete_job(
     return Response(status_code=status.HTTP_204_NO_CONTENT)
 
 
+@router.post("/{job_id}/retry", response_model=JobResponse)
+async def retry_job(
+    job_id: int,
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    """Re-run a failed or stuck job: reset to pending, clear error, relaunch crew."""
+    result = await db.execute(
+        select(Job).where(Job.id == job_id, Job.user_id == current_user.id)
+    )
+    job = result.scalar_one_or_none()
+    if not job:
+        raise HTTPException(status_code=404, detail="Job not found.")
+    if job.status == "completed":
+        raise HTTPException(status_code=409, detail="Completed jobs cannot be retried.")
+    if job.status == "processing":
+        raise HTTPException(status_code=409, detail="Job is already running.")
+
+    # Re-fetch GitHub profile (optional; FK is SET NULL so it may be None already).
+    github_profile = None
+    if job.github_profile_id is not None:
+        gh_result = await db.execute(
+            select(GithubProfile).where(
+                GithubProfile.id == job.github_profile_id,
+                GithubProfile.user_id == current_user.id,
+            )
+        )
+        github_profile = gh_result.scalar_one_or_none()  # gone → run without GitHub
+
+    # Resume must still exist.
+    resume_result = await db.execute(
+        select(Resume).where(Resume.id == job.resume_id, Resume.user_id == current_user.id)
+    )
+    resume = resume_result.scalar_one_or_none()
+    if not resume:
+        raise HTTPException(
+            status_code=409,
+            detail="The resume for this job no longer exists. Cannot retry.",
+        )
+
+    await db.execute(
+        update(Job).where(Job.id == job_id).values(status="pending", error_message=None)
+    )
+    await db.commit()
+    await db.refresh(job)
+
+    # Pre-seed progress so a re-attaching SSE stream doesn't see the stale
+    # terminal 'failed' step and close instantly.
+    progress_store[job_id] = {"current_step": "pending", "error": None}
+
+    asyncio.create_task(
+        run_crew_for_job(
+            job_id=job.id, user_id=current_user.id, user_email=current_user.email,
+            github_profile=github_profile, resume_gcs_path=resume.gcs_path,
+            job_url=job.linkedin_job_url,
+        )
+    )
+    return job
+
+
 @router.get("/{job_id}/progress")
 async def stream_job_progress(
     job_id: int, current_user: User = Depends(get_current_user),

backend/app/cli.py

@@ -106,11 +106,15 @@ def api() -> None:
     """Start the FastAPI app with autoreload (development)."""
     import uvicorn
 
+    # uvicorn excludes whole directory trees only when given absolute paths
+    # (it matches them against each changed file's absolute parents). Relative
+    # globs like ".venv/*" do NOT work. "*.log" still matches by filename.
     uvicorn.run(
         "app.main:app",
         host=DEFAULT_API_HOST,
         port=int(DEFAULT_API_PORT),
         reload=True,
+        reload_excludes=[str(BACKEND_DIR / ".venv"), str(BACKEND_DIR / "logs"), "*.log"],
     )
 
 
@@ -225,6 +229,10 @@ def dev() -> None:
                 [
                     sys.executable, "-m", "uvicorn", "app.main:app",
                     "--host", DEFAULT_API_HOST, "--port", DEFAULT_API_PORT, "--reload",
+                    # Absolute dir paths so uvicorn excludes the whole tree (see api()).
+                    "--reload-exclude", str(BACKEND_DIR / ".venv"),
+                    "--reload-exclude", str(BACKEND_DIR / "logs"),
+                    "--reload-exclude", "*.log",
                 ],
                 BACKEND_DIR,
             )

backend/app/main.py

@@ -5,22 +5,48 @@
 """
 
 from contextlib import asynccontextmanager
+from datetime import datetime, timedelta
 
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
+from sqlalchemy import update
 
 from app.core.config import get_settings
 from app.core.database import engine, Base
+from app.core.logging import get_logger
 from app.api.v1 import auth, github, resumes, jobs
 
 settings = get_settings()
+logger = get_logger(__name__)
 
 
 @asynccontextmanager
 async def lifespan(app: FastAPI):
-    """Create database tables on startup and dispose the engine on shutdown."""
+    """Create tables; best-effort fail jobs stuck >3h; dispose engine on shutdown."""
     async with engine.begin() as conn:
         await conn.run_sync(Base.metadata.create_all)
+
+        # Best-effort: fail jobs left pending/processing for >3h (orphaned by a
+        # prior restart). Reuses THIS connection — no new session/checkout — and
+        # swallows all errors so it can never abort startup or kill the proxy.
+        try:
+            from app.models import Job
+            cutoff = datetime.utcnow() - timedelta(hours=3)
+            await conn.execute(
+                update(Job)
+                .where(
+                    Job.status.in_(("pending", "processing")),
+                    Job.created_at < cutoff,
+                )
+                .values(
+                    status="failed",
+                    error_message="Job was stuck in progress for over 3 hours "
+                                  "(likely a server restart). Please retry.",
+                )
+            )
+        except Exception as exc:  # noqa: BLE001 — never let cleanup break startup
+            logger.warning("Startup stuck-job cleanup skipped: %s", exc)
+
     yield
     await engine.dispose()
 

backend/app/schemas/__init__.py

@@ -79,6 +79,7 @@ class TailorJobRequest(BaseModel):
 
 class JobResponse(BaseModel):
     id: int
+    github_profile_id: Optional[int] = None
     linkedin_job_url: str
     job_name: Optional[str] = None
     company_name: Optional[str] = None

frontend/src/components/TailoredResumeList.jsx

@@ -1,11 +1,25 @@
 import { useState } from 'react';
 import { useNavigate } from 'react-router-dom';
-import { FiEye, FiDownload, FiBookOpen, FiTrash2 } from 'react-icons/fi';
+import { FiEye, FiDownload, FiBookOpen, FiTrash2, FiRefreshCw } from 'react-icons/fi';
 import api from '../services/api';
 
-export default function TailoredResumeList({ jobs, onPreviewResume, onRefresh }) {
+export default function TailoredResumeList({ jobs, onPreviewResume, onRefresh, onRetry }) {
   const navigate = useNavigate();
   const [deletingId, setDeletingId] = useState(null);
+  const [retryingId, setRetryingId] = useState(null);
+
+  const handleRetry = async (job) => {
+    setRetryingId(job.id);
+    try {
+      const res = await api.post(`/jobs/${job.id}/retry`);
+      onRetry?.(res.data);
+    } catch (err) {
+      console.error('Retry failed:', err);
+      alert(err.response?.data?.detail || 'Failed to retry. Please try again.');
+    } finally {
+      setRetryingId(null);
+    }
+  };
 
   const handleDelete = async (job) => {
     const label = job.job_name
@@ -84,6 +98,18 @@ export default function TailoredResumeList({ jobs, onPreviewResume, onRefresh })
                 </button>
               </>
             )}
+            {(job.status === 'failed' || job.status === 'pending') && (
+              <button
+                className="btn btn-secondary btn-sm"
+                onClick={() => handleRetry(job)}
+                disabled={retryingId === job.id}
+                title="Retry"
+              >
+                {retryingId === job.id
+                  ? <span className="spinner" style={{ width: 14, height: 14 }} />
+                  : <><FiRefreshCw size={14} /> Retry</>}
+              </button>
+            )}
             {job.status !== 'pending' && job.status !== 'processing' && (
               <button
                 className="btn btn-ghost btn-icon btn-sm"

frontend/src/pages/HomePage.jsx

@@ -39,6 +39,12 @@ export default function HomePage() {
     setJobs((prev) => [job, ...prev]);
   };
 
+  const handleRetry = (job) => {
+    setActiveJobId(job.id);
+    setActiveJobHasGithub(job.github_profile_id != null);
+    fetchData();
+  };
+
   const handleComplete = useCallback(async () => {
     try {
       const res = await api.get(`/jobs/${activeJobId}`);
@@ -64,7 +70,7 @@ export default function HomePage() {
           <h2 style={{ fontSize: 'var(--font-size-xl)', marginBottom: 'var(--space-4)', display: 'flex', alignItems: 'center', gap: 'var(--space-2)' }}>
             <FiActivity size={20} color="var(--accent-secondary)" /> Previously Tailored Resumes
           </h2>
-          <TailoredResumeList jobs={jobs} onPreviewResume={(id) => setPreviewJobId(id)} onRefresh={fetchData} />
+          <TailoredResumeList jobs={jobs} onPreviewResume={(id) => setPreviewJobId(id)} onRefresh={fetchData} onRetry={handleRetry} />
         </section>
 
         <section style={{ marginBottom: 'var(--space-10)' }}>