diff --git a/AI_API.md b/AI_API.md
index fae9b7a..53ae4df 100644
--- a/AI_API.md
+++ b/AI_API.md
@@ -18,6 +18,12 @@ All requests/responses are JSON. Send `Accept: application/json` (Laravel
 defaults to JSON error responses for API routes regardless, but set it to
 be safe on validation-error bodies).
 
+**If you're an agent running on the machine this app is deployed to**, you
+don't need to build HTTP requests by hand — use the CLI wrappers in
+`python_scripts/` (`sites.py`, `accounts.py`, `posts.py`; see
+`python_scripts/README.md`). They default to `http://localhost:8001` and
+print the same JSON shapes documented below.
+
 ## Mental model
 
 ```
diff --git a/python_scripts/README.md b/python_scripts/README.md
new file mode 100644
index 0000000..544f476
--- /dev/null
+++ b/python_scripts/README.md
@@ -0,0 +1,47 @@
+# Muninn API scripts
+
+Thin Python CLI wrappers around this app's JSON API (see `../AI_API.md` for
+the full API reference). Meant to be called directly by an AI agent on the
+machine where this app is reachable — no setup beyond Python 3's standard
+library (no third-party dependencies).
+
+Default base URL: `http://localhost:8001`. Override with the
+`MUNINN_API_URL` environment variable if the app runs elsewhere or on a
+different port.
+
+Every command prints the API's JSON response to stdout and exits `0` on
+success. On an HTTP error or if the server can't be reached, it prints a
+JSON error object to stderr and exits `1`.
+
+## Sites (read-only)
+
+```
+python3 sites.py list [--name NAME]
+python3 sites.py get SITE_ID
+```
+
+## Social Accounts (read-only)
+
+```
+python3 accounts.py list [--network NETWORK] [--type TYPE]
+python3 accounts.py get ACCOUNT_ID
+```
+
+## Posts (full CRUD)
+
+```
+python3 posts.py list [--status STATUS] [--site-id ID] [--social-account-id ID] [--due]
+python3 posts.py get POST_ID
+python3 posts.py create --site-id ID --type TYPE --body BODY [--social-account-id ID] [--title TITLE] [--status STATUS] [--scheduled-at DATETIME]
+python3 posts.py update POST_ID [--site-id ID] [--social-account-id ID] [--type TYPE] [--title TITLE] [--body BODY] [--status STATUS] [--scheduled-at DATETIME] [--published-url URL]
+python3 posts.py delete POST_ID
+```
+
+`--due` finds posts that are `scheduled` and past their `scheduled_at` —
+the query to run to discover what needs publishing right now.
+
+Publishing itself is out of scope for these scripts (they only wrap the
+HTTP API). See `../AI_API.md` for how to actually publish a post
+(`php artisan posts:publish` for WordPress, the `sidecars/reddit-post`
+sidecar for Reddit, or your own method for anything else — then report
+back with `posts.py update POST_ID --status ...`).
diff --git a/python_scripts/__pycache__/_client.cpython-310.pyc b/python_scripts/__pycache__/_client.cpython-310.pyc
new file mode 100644
index 0000000..bc3c7a8
Binary files /dev/null and b/python_scripts/__pycache__/_client.cpython-310.pyc differ
diff --git a/python_scripts/_client.py b/python_scripts/_client.py
new file mode 100755
index 0000000..649bbc9
--- /dev/null
+++ b/python_scripts/_client.py
@@ -0,0 +1,62 @@
+"""Shared HTTP helper for the Muninn API scripts.
+
+Not meant to be run directly - imported by sites.py, accounts.py, posts.py.
+"""
+
+import json
+import os
+import sys
+import urllib.error
+import urllib.parse
+import urllib.request
+
+DEFAULT_BASE_URL = "http://localhost:8001"
+
+
+def base_url() -> str:
+    return os.environ.get("MUNINN_API_URL", DEFAULT_BASE_URL).rstrip("/")
+
+
+def request(method: str, path: str, params: dict | None = None, body: dict | None = None) -> None:
+    """Call the API and print the JSON response to stdout.
+
+    Exits 0 on a 2xx response. On an HTTP error or network failure, prints
+    a JSON error object to stderr and exits 1.
+    """
+    url = f"{base_url()}{path}"
+
+    if params:
+        query = "&".join(
+            f"{key}={urllib.parse.quote(str(value))}"
+            for key, value in params.items()
+            if value is not None
+        )
+        if query:
+            url = f"{url}?{query}"
+
+    data = json.dumps(body).encode("utf-8") if body is not None else None
+    headers = {"Accept": "application/json"}
+    if data is not None:
+        headers["Content-Type"] = "application/json"
+
+    req = urllib.request.Request(url, data=data, method=method, headers=headers)
+
+    try:
+        with urllib.request.urlopen(req) as response:
+            payload = response.read().decode("utf-8")
+    except urllib.error.HTTPError as error:
+        body_text = error.read().decode("utf-8")
+        try:
+            parsed = json.loads(body_text)
+        except json.JSONDecodeError:
+            parsed = {"error": body_text}
+        print(json.dumps(parsed, indent=2), file=sys.stderr)
+        sys.exit(1)
+    except urllib.error.URLError as error:
+        print(json.dumps({"error": f"could not reach {url}: {error.reason}"}, indent=2), file=sys.stderr)
+        sys.exit(1)
+
+    if payload:
+        parsed = json.loads(payload)
+        print(json.dumps(parsed, indent=2))
+    sys.exit(0)
diff --git a/python_scripts/accounts.py b/python_scripts/accounts.py
new file mode 100755
index 0000000..5465f70
--- /dev/null
+++ b/python_scripts/accounts.py
@@ -0,0 +1,40 @@
+#!/usr/bin/env python3
+"""Read Social Accounts from the Muninn API.
+
+Usage:
+    python3 accounts.py list [--network NETWORK] [--type TYPE]
+    python3 accounts.py get ACCOUNT_ID
+
+Base URL defaults to http://localhost:8001, override with the
+MUNINN_API_URL environment variable.
+"""
+
+import argparse
+
+import _client
+
+NETWORKS = ["twitter", "instagram", "facebook", "reddit", "tiktok", "linkedin", "youtube", "other"]
+TYPES = ["post", "comment", "like"]
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="Social Accounts API client")
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    list_parser = subparsers.add_parser("list", help="List all social accounts")
+    list_parser.add_argument("--network", choices=NETWORKS, help="Exact match filter on network")
+    list_parser.add_argument("--type", choices=TYPES, help="Exact match filter on type")
+
+    get_parser = subparsers.add_parser("get", help="Get a single social account")
+    get_parser.add_argument("account_id", type=int)
+
+    args = parser.parse_args()
+
+    if args.command == "list":
+        _client.request("GET", "/api/accounts", params={"network": args.network, "type": args.type})
+    elif args.command == "get":
+        _client.request("GET", f"/api/accounts/{args.account_id}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/python_scripts/posts.py b/python_scripts/posts.py
new file mode 100755
index 0000000..23cd8dc
--- /dev/null
+++ b/python_scripts/posts.py
@@ -0,0 +1,113 @@
+#!/usr/bin/env python3
+"""Read and write Posts on the Muninn API.
+
+Usage:
+    python3 posts.py list [--status STATUS] [--site-id ID] [--social-account-id ID] [--due]
+    python3 posts.py get POST_ID
+    python3 posts.py create --site-id ID --type TYPE --body BODY
+                             [--social-account-id ID] [--title TITLE]
+                             [--status STATUS] [--scheduled-at DATETIME]
+    python3 posts.py update POST_ID [--site-id ID] [--social-account-id ID]
+                             [--type TYPE] [--title TITLE] [--body BODY]
+                             [--status STATUS] [--scheduled-at DATETIME]
+                             [--published-url URL]
+    python3 posts.py delete POST_ID
+
+Base URL defaults to http://localhost:8001, override with the
+MUNINN_API_URL environment variable.
+"""
+
+import argparse
+
+import _client
+
+TYPES = ["twitter", "instagram", "facebook", "reddit", "tiktok", "linkedin", "youtube", "blog", "other"]
+STATUSES = ["draft", "scheduled", "published", "failed"]
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="Posts API client")
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    list_parser = subparsers.add_parser("list", help="List posts")
+    list_parser.add_argument("--status", choices=STATUSES)
+    list_parser.add_argument("--site-id", type=int)
+    list_parser.add_argument("--social-account-id", type=int)
+    list_parser.add_argument(
+        "--due", action="store_true", help="Only posts that are scheduled and due now"
+    )
+
+    get_parser = subparsers.add_parser("get", help="Get a single post")
+    get_parser.add_argument("post_id", type=int)
+
+    create_parser = subparsers.add_parser("create", help="Create a post")
+    create_parser.add_argument("--site-id", type=int, required=True)
+    create_parser.add_argument("--type", choices=TYPES, required=True)
+    create_parser.add_argument("--body", required=True)
+    create_parser.add_argument("--social-account-id", type=int)
+    create_parser.add_argument("--title")
+    create_parser.add_argument("--status", choices=STATUSES)
+    create_parser.add_argument("--scheduled-at", help="ISO 8601 datetime, e.g. 2026-07-20T09:00:00")
+
+    update_parser = subparsers.add_parser("update", help="Update a post")
+    update_parser.add_argument("post_id", type=int)
+    update_parser.add_argument("--site-id", type=int)
+    update_parser.add_argument("--social-account-id", type=int)
+    update_parser.add_argument("--type", choices=TYPES)
+    update_parser.add_argument("--title")
+    update_parser.add_argument("--body")
+    update_parser.add_argument("--status", choices=STATUSES)
+    update_parser.add_argument("--scheduled-at", help="ISO 8601 datetime")
+    update_parser.add_argument("--published-url")
+
+    delete_parser = subparsers.add_parser("delete", help="Delete a post")
+    delete_parser.add_argument("post_id", type=int)
+
+    args = parser.parse_args()
+
+    if args.command == "list":
+        params = {
+            "status": args.status,
+            "site_id": args.site_id,
+            "social_account_id": args.social_account_id,
+        }
+        if args.due:
+            params["due"] = 1
+        _client.request("GET", "/api/posts", params=params)
+
+    elif args.command == "get":
+        _client.request("GET", f"/api/posts/{args.post_id}")
+
+    elif args.command == "create":
+        body = {
+            "site_id": args.site_id,
+            "social_account_id": args.social_account_id,
+            "type": args.type,
+            "title": args.title,
+            "body": args.body,
+            "status": args.status,
+            "scheduled_at": args.scheduled_at,
+        }
+        body = {key: value for key, value in body.items() if value is not None}
+        _client.request("POST", "/api/posts", body=body)
+
+    elif args.command == "update":
+        body = {
+            "site_id": args.site_id,
+            "social_account_id": args.social_account_id,
+            "type": args.type,
+            "title": args.title,
+            "body": args.body,
+            "status": args.status,
+            "scheduled_at": args.scheduled_at,
+            "published_url": args.published_url,
+        }
+        body = {key: value for key, value in body.items() if value is not None}
+        _client.request("PATCH", f"/api/posts/{args.post_id}", body=body)
+
+    elif args.command == "delete":
+        _client.request("DELETE", f"/api/posts/{args.post_id}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/python_scripts/sites.py b/python_scripts/sites.py
new file mode 100755
index 0000000..800f388
--- /dev/null
+++ b/python_scripts/sites.py
@@ -0,0 +1,36 @@
+#!/usr/bin/env python3
+"""Read Sites from the Muninn API.
+
+Usage:
+    python3 sites.py list [--name NAME]
+    python3 sites.py get SITE_ID
+
+Base URL defaults to http://localhost:8001, override with the
+MUNINN_API_URL environment variable.
+"""
+
+import argparse
+
+import _client
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="Sites API client")
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    list_parser = subparsers.add_parser("list", help="List all sites")
+    list_parser.add_argument("--name", help="Partial match filter on name")
+
+    get_parser = subparsers.add_parser("get", help="Get a single site")
+    get_parser.add_argument("site_id", type=int)
+
+    args = parser.parse_args()
+
+    if args.command == "list":
+        _client.request("GET", "/api/sites", params={"name": args.name})
+    elif args.command == "get":
+        _client.request("GET", f"/api/sites/{args.site_id}")
+
+
+if __name__ == "__main__":
+    main()