diff --git a/repub/jobs.py b/repub/jobs.py
index f339195..5d3ed7f 100644
--- a/repub/jobs.py
+++ b/repub/jobs.py
@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import json
+import math
 import os
 import signal
 import subprocess
@@ -10,7 +11,7 @@ import time
 from dataclasses import dataclass
 from datetime import UTC, datetime, timedelta
 from pathlib import Path
-from typing import Callable, TextIO, cast
+from typing import Callable, TextIO, TypedDict, cast
 
 from apscheduler.schedulers.background import BackgroundScheduler
 from apscheduler.triggers.cron import CronTrigger
@@ -30,6 +31,7 @@ from repub.model import (
 SCHEDULER_JOB_PREFIX = "job-"
 POLL_JOB_ID = "runtime-poll-workers"
 SYNC_JOB_ID = "runtime-sync-jobs"
+COMPLETED_EXECUTION_PAGE_SIZE = 20
 
 
 @dataclass(frozen=True)
@@ -102,6 +104,17 @@ class ExecutionLogView:
     error_message: str | None = None
 
 
+class RunsView(TypedDict):
+    running: tuple[dict[str, object], ...]
+    queued: tuple[dict[str, object], ...]
+    upcoming: tuple[dict[str, object], ...]
+    completed: tuple[dict[str, object], ...]
+    completed_page: int
+    completed_page_size: int
+    completed_total_count: int
+    completed_total_pages: int
+
+
 class JobRuntime:
     def __init__(
         self,
@@ -647,10 +660,15 @@ class JobRuntime:
 
 
 def load_runs_view(
-    *, log_dir: str | Path, now: datetime | None = None
-) -> dict[str, tuple[dict[str, object], ...]]:
+    *,
+    log_dir: str | Path,
+    now: datetime | None = None,
+    completed_page: int = 1,
+    completed_page_size: int = COMPLETED_EXECUTION_PAGE_SIZE,
+) -> RunsView:
     reference_time = now or datetime.now(UTC)
     resolved_log_dir = Path(log_dir)
+    sanitized_page_size = max(1, completed_page_size)
     with database.connection_context():
         execution_primary_key = getattr(JobExecution, "_meta").primary_key
         jobs = tuple(Job.select(Job, Source).join(Source).order_by(Source.name.asc()))
@@ -668,7 +686,7 @@ def load_runs_view(
             .where(JobExecution.running_status == JobExecutionStatus.RUNNING)
             .order_by(JobExecution.started_at.desc())
         )
-        completed_executions = tuple(
+        completed_query = (
             JobExecution.select(JobExecution, Job, Source)
             .join(Job)
             .join(Source)
@@ -682,7 +700,14 @@ def load_runs_view(
                 )
             )
             .order_by(JobExecution.ended_at.desc())
-            .limit(20)
+        )
+        completed_total_count = completed_query.count()
+        completed_total_pages = max(
+            1, math.ceil(completed_total_count / sanitized_page_size)
+        )
+        sanitized_completed_page = min(max(1, completed_page), completed_total_pages)
+        completed_executions = tuple(
+            completed_query.paginate(sanitized_completed_page, sanitized_page_size)
         )
 
         running_by_job = {
@@ -725,9 +750,49 @@ def load_runs_view(
             _project_completed_execution(execution, resolved_log_dir, reference_time)
             for execution in completed_executions
         ),
+        "completed_page": sanitized_completed_page,
+        "completed_page_size": sanitized_page_size,
+        "completed_total_count": completed_total_count,
+        "completed_total_pages": completed_total_pages,
     }
 
 
+def clear_completed_executions(*, log_dir: str | Path) -> int:
+    resolved_log_dir = Path(log_dir)
+    with database.connection_context():
+        execution_primary_key = getattr(JobExecution, "_meta").primary_key
+        completed_executions = tuple(
+            JobExecution.select(JobExecution, Job)
+            .join(Job)
+            .where(
+                JobExecution.running_status.in_(
+                    (
+                        JobExecutionStatus.SUCCEEDED,
+                        JobExecutionStatus.FAILED,
+                        JobExecutionStatus.CANCELED,
+                    )
+                )
+            )
+        )
+        if not completed_executions:
+            return 0
+
+        for execution in completed_executions:
+            job = cast(Job, execution.job)
+            prefix = f"job-{_job_id(job)}-execution-{_execution_id(execution)}"
+            for artifact_path in resolved_log_dir.glob(f"{prefix}.*"):
+                artifact_path.unlink(missing_ok=True)
+
+        execution_ids = tuple(
+            _execution_id(execution) for execution in completed_executions
+        )
+        return (
+            JobExecution.delete()
+            .where(execution_primary_key.in_(execution_ids))
+            .execute()
+        )
+
+
 def load_dashboard_view(
     *, log_dir: str | Path, now: datetime | None = None
 ) -> dict[str, object]:
diff --git a/repub/pages/runs.py b/repub/pages/runs.py
index 8ed592a..cecd089 100644
--- a/repub/pages/runs.py
+++ b/repub/pages/runs.py
@@ -242,12 +242,129 @@ def _completed_row(execution: Mapping[str, object]) -> tuple[Node, ...]:
     )
 
 
+def _completed_page_href(page: int) -> str:
+    return f"/runs?completed_page={page}"
+
+
+def _completed_history_pagination(
+    *,
+    completed_page: int,
+    completed_page_size: int,
+    completed_total_count: int,
+    completed_total_pages: int,
+) -> Renderable | None:
+    if completed_total_count <= completed_page_size:
+        return None
+
+    start_result = ((completed_page - 1) * completed_page_size) + 1
+    end_result = min(completed_total_count, completed_page * completed_page_size)
+    link_class = (
+        "relative inline-flex items-center px-4 py-2 text-sm font-semibold text-slate-700 "
+        "ring-1 ring-inset ring-slate-200 hover:bg-stone-50"
+    )
+
+    return h.div(
+        class_="flex items-center justify-between border-t border-slate-200 bg-white px-4 py-3 sm:px-6"
+    )[
+        h.div(class_="flex flex-1 justify-between sm:hidden")[
+            h.a(
+                href=_completed_page_href(max(1, completed_page - 1)),
+                class_="relative inline-flex items-center rounded-xl border border-slate-200 bg-white px-4 py-2 text-sm font-medium text-slate-700 hover:bg-stone-50",
+            )["Previous"],
+            h.a(
+                href=_completed_page_href(
+                    min(completed_total_pages, completed_page + 1)
+                ),
+                class_="relative ml-3 inline-flex items-center rounded-xl border border-slate-200 bg-white px-4 py-2 text-sm font-medium text-slate-700 hover:bg-stone-50",
+            )["Next"],
+        ],
+        h.div(class_="hidden sm:flex sm:flex-1 sm:items-center sm:justify-between")[
+            h.p(class_="text-sm text-slate-600")[
+                "Showing ",
+                h.span(class_="font-medium text-slate-950")[str(start_result)],
+                " to ",
+                h.span(class_="font-medium text-slate-950")[str(end_result)],
+                " of ",
+                h.span(class_="font-medium text-slate-950")[str(completed_total_count)],
+                " results",
+            ],
+            h.nav(
+                aria_label="Completed execution pagination",
+                class_="isolate inline-flex -space-x-px rounded-xl shadow-xs",
+            )[
+                (
+                    h.a(
+                        href=_completed_page_href(page_number),
+                        aria_current=(
+                            "page" if page_number == completed_page else None
+                        ),
+                        class_=(
+                            "relative z-10 inline-flex items-center bg-amber-500 px-4 py-2 text-sm font-semibold text-slate-950"
+                            if page_number == completed_page
+                            else link_class
+                        ),
+                    )[str(page_number)]
+                    for page_number in range(1, completed_total_pages + 1)
+                )
+            ],
+        ],
+    ]
+
+
+def _completed_history_section(
+    *,
+    completed_rows: tuple[tuple[Node, ...], ...],
+    completed_page: int,
+    completed_page_size: int,
+    completed_total_count: int,
+    completed_total_pages: int,
+) -> Renderable:
+    pagination = _completed_history_pagination(
+        completed_page=completed_page,
+        completed_page_size=completed_page_size,
+        completed_total_count=completed_total_count,
+        completed_total_pages=completed_total_pages,
+    )
+    return h.section[
+        table_section(
+            eyebrow="History",
+            title="Completed job executions",
+            empty_message="No job executions have completed yet.",
+            headers=(
+                "#",
+                "Source",
+                "Ended",
+                "State",
+                "Summary",
+                "Log",
+            ),
+            rows=completed_rows,
+            first_header_class="w-px py-2.5 pr-2 pl-3 text-left text-xs font-semibold uppercase tracking-[0.18em] whitespace-nowrap text-slate-500 sm:pl-3",
+            first_cell_class="w-px py-3 pr-2 pl-3 text-sm font-medium text-slate-950 sm:pl-3",
+            actions=(
+                action_button(
+                    label="Clear history",
+                    tone="danger",
+                    post_path="/actions/completed-executions/clear",
+                )
+                if completed_total_count > 0
+                else None
+            ),
+        ),
+        pagination,
+    ]
+
+
 def runs_page(
     *,
     running_executions: tuple[Mapping[str, object], ...] | None = None,
     queued_executions: tuple[Mapping[str, object], ...] | None = None,
     upcoming_jobs: tuple[Mapping[str, object], ...] | None = None,
     completed_executions: tuple[Mapping[str, object], ...] | None = None,
+    completed_page: int = 1,
+    completed_page_size: int = 20,
+    completed_total_count: int | None = None,
+    completed_total_pages: int | None = None,
     source_count: int = 0,
 ) -> Renderable:
     running_items = running_executions or ()
@@ -262,6 +379,12 @@ def runs_page(
     )
     upcoming_rows = tuple(_upcoming_row(job) for job in upcoming_items)
     completed_rows = tuple(_completed_row(execution) for execution in completed_items)
+    resolved_completed_total_count = (
+        len(completed_items) if completed_total_count is None else completed_total_count
+    )
+    resolved_completed_total_pages = (
+        1 if completed_total_pages is None else completed_total_pages
+    )
 
     return page_shell(
         current_path="/runs",
@@ -302,21 +425,12 @@ def runs_page(
                 ),
                 rows=upcoming_rows,
             ),
-            table_section(
-                eyebrow="History",
-                title="Completed job executions",
-                empty_message="No job executions have completed yet.",
-                headers=(
-                    "#",
-                    "Source",
-                    "Ended",
-                    "State",
-                    "Summary",
-                    "Log",
-                ),
-                rows=completed_rows,
-                first_header_class="w-px py-2.5 pr-2 pl-3 text-left text-xs font-semibold uppercase tracking-[0.18em] whitespace-nowrap text-slate-500 sm:pl-3",
-                first_cell_class="w-px py-3 pr-2 pl-3 text-sm font-medium text-slate-950 sm:pl-3",
+            _completed_history_section(
+                completed_rows=completed_rows,
+                completed_page=completed_page,
+                completed_page_size=completed_page_size,
+                completed_total_count=resolved_completed_total_count,
+                completed_total_pages=resolved_completed_total_pages,
             ),
             h.script[
                 """
diff --git a/repub/pages/sources.py b/repub/pages/sources.py
index 93c625f..62d1e9a 100644
--- a/repub/pages/sources.py
+++ b/repub/pages/sources.py
@@ -75,13 +75,11 @@ def _source_row(source: Mapping[str, object]) -> tuple[Node, ...]:
                 label=str(source["state"]),
                 tone=str(source["state_tone"]),
             ),
-            h.p(class_="mt-2 text-xs text-slate-500")[str(source["last_run"])],
         ],
-        h.div(class_="flex flex-wrap items-center gap-2")[
+        h.div(class_="flex flex-nowrap items-center gap-3 whitespace-nowrap")[
             inline_link(
                 href=f"/sources/{source['slug']}/edit", label="Edit", tone="amber"
             ),
-            inline_link(href="/runs", label="View runs"),
             action_button(
                 label="Delete",
                 tone="danger",
diff --git a/repub/static/app.css b/repub/static/app.css
index 3d3aa98..75508d6 100644
--- a/repub/static/app.css
+++ b/repub/static/app.css
@@ -251,6 +251,12 @@
   .top-0 {
     top: calc(var(--spacing) * 0);
   }
+  .isolate {
+    isolation: isolate;
+  }
+  .z-10 {
+    z-index: 10;
+  }
   .container {
     width: 100%;
     @media (width >= 40rem) {
@@ -299,6 +305,9 @@
   .mb-3 {
     margin-bottom: calc(var(--spacing) * 3);
   }
+  .ml-3 {
+    margin-left: calc(var(--spacing) * 3);
+  }
   .block {
     display: block;
   }
@@ -393,6 +402,9 @@
   .min-w-\[70rem\] {
     min-width: 70rem;
   }
+  .flex-1 {
+    flex: 1;
+  }
   .shrink-0 {
     flex-shrink: 0;
   }
@@ -475,6 +487,13 @@
       margin-block-end: calc(calc(var(--spacing) * 6) * calc(1 - var(--tw-space-y-reverse)));
     }
   }
+  .-space-x-px {
+    :where(& > :not(:last-child)) {
+      --tw-space-x-reverse: 0;
+      margin-inline-start: calc(-1px * var(--tw-space-x-reverse));
+      margin-inline-end: calc(-1px * calc(1 - var(--tw-space-x-reverse)));
+    }
+  }
   .divide-y {
     :where(& > :not(:last-child)) {
       --tw-divide-y-reverse: 0;
@@ -912,6 +931,9 @@
     transition-timing-function: var(--tw-ease, var(--default-transition-timing-function));
     transition-duration: var(--tw-duration, var(--default-transition-duration));
   }
+  .ring-inset {
+    --tw-ring-inset: inset;
+  }
   .placeholder\:text-slate-400 {
     &::placeholder {
       color: var(--color-slate-400);
@@ -962,6 +984,13 @@
       }
     }
   }
+  .hover\:bg-stone-50 {
+    &:hover {
+      @media (hover: hover) {
+        background-color: var(--color-stone-50);
+      }
+    }
+  }
   .hover\:bg-stone-200 {
     &:hover {
       @media (hover: hover) {
@@ -1054,6 +1083,21 @@
       background-color: var(--color-slate-200);
     }
   }
+  .sm\:flex {
+    @media (width >= 40rem) {
+      display: flex;
+    }
+  }
+  .sm\:hidden {
+    @media (width >= 40rem) {
+      display: none;
+    }
+  }
+  .sm\:flex-1 {
+    @media (width >= 40rem) {
+      flex: 1;
+    }
+  }
   .sm\:grid-cols-2 {
     @media (width >= 40rem) {
       grid-template-columns: repeat(2, minmax(0, 1fr));
@@ -1064,6 +1108,11 @@
       flex-direction: row;
     }
   }
+  .sm\:items-center {
+    @media (width >= 40rem) {
+      align-items: center;
+    }
+  }
   .sm\:items-end {
     @media (width >= 40rem) {
       align-items: flex-end;
@@ -1084,6 +1133,11 @@
       padding-inline: calc(var(--spacing) * 4);
     }
   }
+  .sm\:px-6 {
+    @media (width >= 40rem) {
+      padding-inline: calc(var(--spacing) * 6);
+    }
+  }
   .sm\:pl-2\.5 {
     @media (width >= 40rem) {
       padding-left: calc(var(--spacing) * 2.5);
@@ -1188,6 +1242,11 @@
   inherits: false;
   initial-value: 0;
 }
+@property --tw-space-x-reverse {
+  syntax: "*";
+  inherits: false;
+  initial-value: 0;
+}
 @property --tw-divide-y-reverse {
   syntax: "*";
   inherits: false;
@@ -1418,6 +1477,7 @@
       --tw-translate-y: 0;
       --tw-translate-z: 0;
       --tw-space-y-reverse: 0;
+      --tw-space-x-reverse: 0;
       --tw-divide-y-reverse: 0;
       --tw-border-style: solid;
       --tw-gradient-position: initial;
diff --git a/repub/web.py b/repub/web.py
index 56c75b9..5f0c065 100644
--- a/repub/web.py
+++ b/repub/web.py
@@ -13,11 +13,20 @@ from datastar_py.quart import DatastarResponse, read_signals
 from datastar_py.sse import DatastarEvent
 from htpy import Renderable
 from peewee import IntegrityError
-from quart import Quart, Response, request, send_from_directory, url_for
+from quart import (
+    Quart,
+    Response,
+    has_request_context,
+    request,
+    send_from_directory,
+    url_for,
+)
 
 from repub.datastar import RefreshBroker, render_stream
 from repub.jobs import (
+    COMPLETED_EXECUTION_PAGE_SIZE,
     JobRuntime,
+    clear_completed_executions,
     load_dashboard_view,
     load_execution_log_view,
     load_runs_view,
@@ -329,6 +338,12 @@ def create_app(*, dev_mode: bool = False) -> Quart:
         get_job_runtime(app).move_queued_execution(execution_id, direction="down")
         return Response(status=204)
 
+    @app.post("/actions/completed-executions/clear")
+    async def clear_completed_executions_action() -> Response:
+        clear_completed_executions(log_dir=app.config["REPUB_LOG_DIR"])
+        trigger_refresh(app)
+        return Response(status=204)
+
     @app.post("/job/<int:job_id>/execution/<int:execution_id>/logs")
     async def logs_patch(job_id: int, execution_id: int) -> DatastarResponse:
         async def render() -> Renderable:
@@ -420,12 +435,25 @@ async def render_runs(app: Quart | None = None) -> Renderable:
     if app is None:
         return runs_page()
 
-    view = load_runs_view(log_dir=app.config["REPUB_LOG_DIR"])
+    completed_page = (
+        max(1, request.args.get("completed_page", 1, type=int) or 1)
+        if has_request_context()
+        else 1
+    )
+    view = load_runs_view(
+        log_dir=app.config["REPUB_LOG_DIR"],
+        completed_page=completed_page,
+        completed_page_size=COMPLETED_EXECUTION_PAGE_SIZE,
+    )
     return runs_page(
         running_executions=cast(tuple[dict[str, object], ...], view["running"]),
         queued_executions=cast(tuple[dict[str, object], ...], view["queued"]),
         upcoming_jobs=cast(tuple[dict[str, object], ...], view["upcoming"]),
         completed_executions=cast(tuple[dict[str, object], ...], view["completed"]),
+        completed_page=cast(int, view["completed_page"]),
+        completed_page_size=cast(int, view["completed_page_size"]),
+        completed_total_count=cast(int, view["completed_total_count"]),
+        completed_total_pages=cast(int, view["completed_total_pages"]),
         source_count=len(load_sources()),
     )
 
diff --git a/tests/test_jobs.py b/tests/test_jobs.py
index 6159c74..45cc42b 100644
--- a/tests/test_jobs.py
+++ b/tests/test_jobs.py
@@ -254,3 +254,51 @@ def test_load_runs_view_running_row_targets_queued_follow_up_cancel(
     assert running_row["cancel_post_path"] == (
         f"/actions/queued-executions/{int(pending_execution.get_id())}/cancel"
     )
+
+
+def test_load_runs_view_paginates_completed_executions_after_20_rows(
+    tmp_path: Path,
+) -> None:
+    initialize_database(tmp_path / "jobs-completed-pagination.db")
+    source = create_source(
+        name="Completed source",
+        slug="completed-source",
+        source_type="feed",
+        notes="",
+        spider_arguments="",
+        enabled=False,
+        cron_minute="*/5",
+        cron_hour="*",
+        cron_day_of_month="*",
+        cron_day_of_week="*",
+        cron_month="*",
+        feed_url="https://example.com/completed.xml",
+    )
+    job = Job.get(Job.source == source)
+    base_time = datetime(2026, 3, 30, 12, 0, tzinfo=UTC)
+    for offset in range(21):
+        JobExecution.create(
+            job=job,
+            running_status=JobExecutionStatus.SUCCEEDED,
+            ended_at=base_time - timedelta(minutes=offset),
+        )
+
+    first_page = load_runs_view(
+        log_dir=tmp_path / "out" / "logs",
+        now=base_time,
+        completed_page=1,
+    )
+    second_page = load_runs_view(
+        log_dir=tmp_path / "out" / "logs",
+        now=base_time,
+        completed_page=2,
+    )
+
+    assert len(first_page["completed"]) == 20
+    assert len(second_page["completed"]) == 1
+    assert first_page["completed_page"] == 1
+    assert second_page["completed_page"] == 2
+    assert first_page["completed_total_pages"] == 2
+    assert second_page["completed_total_pages"] == 2
+    assert first_page["completed_total_count"] == 21
+    assert second_page["completed_total_count"] == 21
diff --git a/tests/test_web.py b/tests/test_web.py
index ad64b38..c309f48 100644
--- a/tests/test_web.py
+++ b/tests/test_web.py
@@ -22,6 +22,7 @@ from repub.model import (
     save_setting,
 )
 from repub.pages.runs import runs_page
+from repub.pages.sources import sources_page
 from repub.web import (
     create_app,
     get_refresh_broker,
@@ -172,6 +173,66 @@ def test_runs_page_renders_combined_running_jobs_table() -> None:
     assert "/actions/queued-executions/42/cancel" in body
 
 
+def test_sources_page_removes_view_runs_action_and_last_run_caption() -> None:
+    body = str(
+        sources_page(
+            sources=(
+                {
+                    "name": "Source one",
+                    "slug": "source-one",
+                    "source_type": "Feed",
+                    "upstream": "https://example.com/feed.xml",
+                    "schedule": "cron: */5 * * * *",
+                    "last_run": "Never run",
+                    "state": "Enabled",
+                    "state_tone": "scheduled",
+                },
+            )
+        )
+    )
+
+    assert ">Edit<" in body
+    assert ">Delete<" in body
+    assert "View runs" not in body
+    assert "Never run" not in body
+
+
+def test_runs_page_renders_clear_completed_button_and_pagination() -> None:
+    completed_executions = tuple(
+        {
+            "source": f"Completed source {index}",
+            "slug": f"completed-source-{index}",
+            "job_id": 7,
+            "execution_id": index,
+            "ended_at": "2 hours ago",
+            "ended_at_iso": "2026-01-15T10:00:00+00:00",
+            "status": "Succeeded",
+            "status_tone": "done",
+            "stats": "1 requests • 1 items • 1 bytes",
+            "summary": "Worker exited successfully",
+            "log_href": f"/job/7/execution/{index}/logs",
+        }
+        for index in range(1, 21)
+    )
+    body = str(
+        runs_page(
+            completed_executions=completed_executions,
+            completed_page=2,
+            completed_page_size=20,
+            completed_total_count=21,
+            completed_total_pages=2,
+        )
+    )
+
+    assert "/actions/completed-executions/clear" in body
+    assert ">Clear history<" in body
+    assert "Showing" in body
+    assert "21" in body
+    assert 'href="/runs?completed_page=1"' in body
+    assert 'href="/runs?completed_page=2"' in body
+    assert 'aria-current="page"' in body
+
+
 def test_root_get_serves_datastar_shim() -> None:
     async def run() -> None:
         client = create_app().test_client()
@@ -1498,6 +1559,66 @@ def test_cancel_queued_execution_action_deletes_pending_row_without_touching_run
     asyncio.run(run())
 
 
+def test_clear_completed_executions_action_removes_history_and_log_artifacts(
+    monkeypatch, tmp_path: Path
+) -> None:
+    db_path = tmp_path / "clear-completed-action.db"
+    log_dir = tmp_path / "out" / "logs"
+    monkeypatch.setenv("REPUBLISHER_DB_PATH", str(db_path))
+
+    async def run() -> None:
+        app = create_app()
+        app.config["REPUB_LOG_DIR"] = log_dir
+        client = app.test_client()
+
+        source = create_source(
+            name="History source",
+            slug="history-source",
+            source_type="feed",
+            notes="",
+            spider_arguments="",
+            enabled=True,
+            cron_minute="*/5",
+            cron_hour="*",
+            cron_day_of_month="*",
+            cron_day_of_week="*",
+            cron_month="*",
+            feed_url="https://example.com/history.xml",
+        )
+        job = Job.get(Job.source == source)
+        completed_execution = JobExecution.create(
+            job=job,
+            running_status=JobExecutionStatus.SUCCEEDED,
+            ended_at=datetime(2026, 3, 30, 12, 0, tzinfo=UTC),
+        )
+        running_execution = JobExecution.create(
+            job=job,
+            running_status=JobExecutionStatus.RUNNING,
+            started_at=datetime(2026, 3, 30, 12, 5, tzinfo=UTC),
+        )
+        log_dir.mkdir(parents=True, exist_ok=True)
+        completed_prefix = (
+            log_dir / f"job-{job.id}-execution-{int(completed_execution.get_id())}"
+        )
+        running_log_path = (
+            log_dir / f"job-{job.id}-execution-{int(running_execution.get_id())}.log"
+        )
+        for suffix in (".log", ".jsonl", ".pygea.log"):
+            completed_prefix.with_suffix(suffix).write_text("history", encoding="utf-8")
+        running_log_path.write_text("running", encoding="utf-8")
+
+        response = await client.post("/actions/completed-executions/clear")
+
+        assert response.status_code == 204
+        assert JobExecution.get_or_none(id=int(completed_execution.get_id())) is None
+        assert JobExecution.get_or_none(id=int(running_execution.get_id())) is not None
+        for suffix in (".log", ".jsonl", ".pygea.log"):
+            assert not completed_prefix.with_suffix(suffix).exists()
+        assert running_log_path.exists()
+
+    asyncio.run(run())
+
+
 def test_move_queued_execution_action_reorders_queue(
     monkeypatch, tmp_path: Path
 ) -> None: