medium-ops

Standalone Medium CLI + 22-tool MCP server. Your IDE drafts the replies. Zero AI API keys.
Stories, responses, claps, feed, profiles, stats, reply engine, MCP server.
One Python install, one binary, MIT licensed. Sibling of
substack-ops.
TL;DR — MCP-native (no API key, one command)
uvx medium-ops mcp install cursor
Your host's LLM (Cursor's, Claude's) does the drafting via the
propose_reply / confirm_reply tools. No ANTHROPIC_API_KEY /
OPENAI_API_KEY needed.
Why a hybrid
Medium exposes three usable surfaces and we use all of them:
- Public RSS (reads, no auth).
medium.com/feed/@{user} returns the
author's ~10 most recent stories with body_html, pubDate, tags,
hero image, and dc:creator. Zero credentials, faster than GraphQL,
stable. Used by default for list_posts / get_post / get_post_content.
Inspired by Portfolio_V2's blog page.
- Dashboard GraphQL (authenticated reads).
medium.com/_/graphql +
medium.com/_/api/* with the sid cookie. Used as a fallback when
you ask for more than ~10 posts, when the post isn't in the RSS window,
or for things RSS can't give you (responses, claps, feed, stats, search).
- Official REST (writes).
api.medium.com/v1/* with an
Integration Token. Supports createPost, createPostInPublication,
getUser, getPublications. That's it.
Force a specific transport with --source rss|graphql|auto on posts list,
posts show, and posts content. The dashboard + GraphQL endpoints are
undocumented and Medium can change them at any time. See
Known gaps.
Setup (dev / from source)
git clone https://github.com/06ketan/medium-ops && cd medium-ops
uv sync
uv sync --extra mcp
uv sync --extra tui
Auth is read from ~/.cursor/mcp.json's mcpServers.medium-ops.env (or
medium-api / medium). Override with env or .env.
uv run medium-ops auth verify
uv run medium-ops quickstart
Command surface
Every write defaults to --dry-run. Flip with --no-dry-run. All writes
land in .cache/audit.jsonl and are dedup-checked against
.cache/actions.db.
Auth (3)
auth verify | Probe both integration token (/me) and sid cookie (GraphQL Viewer). |
auth test | Same but exits non-zero on failure (CI-friendly). |
auth setup | Interactive: paste token / sid / uid / username to .env. |
Read — Stories (5)
posts list [--user] [--limit] | Latest stories by a user (default: self). |
posts show <id_or_url> | Story metadata (title, clap count, response count). |
posts content <id> [--md] | Body HTML (or Markdown with --md). |
posts search <query> [--limit] | Medium-side full-text search. |
| `posts publish -t "..." -f body.md [--pub] [--status draft | public |
Read + Write — Responses (3)
responses list <post_id> [--limit] | Top-level responses table. |
responses tree <post_id> [--out file.json] | Full response + reply tree JSON. |
responses add <post_id> "body" [--parent <r_id>] [--no-dry-run] | Post a response or reply. |
Read + Write — Claps (2)
claps count <post_id> | Total claps. |
claps give <post_id> [--claps N] [--no-dry-run] | Clap 1-50 times. Dedup-protected. |
Read — Discovery + Profile (5)
feed list [--tab home|following|tag-{slug}] [--limit] | Reader feed. |
profile me | Your full profile (GraphQL). |
profile get <username> | Any user's public profile. |
profile stats [--days N] | Per-post views / reads / fans (dashboard scrape). |
profile publications | Publications you can publish to (integration token). |
Reply engine (3)
reply template <post_id> --template thanks | Rule-based replies (no LLM). |
reply bulk <post_id> --out drafts.json | Draft every response to a file. |
reply bulk-send drafts.json [--no-dry-run] | Post only action=approved rows. Dedup-checked. |
Operations + safety (2)
audit search [--kind] [--target] [--status] [--since 7d] | Query the JSONL audit log. |
audit dedup-status | Counts in the dedup SQLite DB. |
MCP server (3)
mcp install <cursor|claude-desktop|claude-code|print> [--dry-run] | Auto-merge config into your host. |
mcp serve | stdio MCP server (22 tools). |
mcp list-tools | Print the tool registry. |
Other (1)
quickstart | Print a quickstart checklist. |
Reply modes
template | YAML keyword rules under src/medium_ops/templates/*.yaml | dry-run default |
bulk | LLM drafts every response to drafts.json. Edit, set action: "approved" | offline review, dedup-checked on send |
bulk-send | Posts only items with action: "approved" | dry-run default; dedup DB prevents dup replies |
MCP propose_reply → confirm_reply | Host LLM drafts, you approve per-item, token-gated | 5-min token TTL, idempotent, no API key |
MCP server
medium-ops mcp install cursor
medium-ops mcp install claude-desktop
medium-ops mcp install claude-code
medium-ops mcp install print
medium-ops mcp serve
medium-ops mcp list-tools
Manual config snippet:
{
"mcpServers": {
"medium-ops": {
"command": "medium-ops",
"args": ["mcp", "serve"]
}
}
}
If the mcp SDK is not installed, the server falls back to a minimal
stdin/stdout JSON-line dispatcher:
echo '{"tool":"list_posts","args":{"limit":3}}' | medium-ops mcp serve
MCP-native draft loop (no API key)
The safety + drafting stack that makes the unattended mode safe:
get_unanswered_responses | Worklist — responses where you haven't replied. |
propose_reply | Dry-run only. Returns a token + payload preview. |
confirm_reply | Posts the staged reply by token. Idempotent via dedup DB. Token TTL 5 min. |
bulk_draft_replies / send_approved_drafts | File-based offline review loop. |
audit_search / dedup_status | Read the audit log + dedup counts. |
LLM strategy
Two layers, both free:
- MCP-native (default). Host LLM drafts via
propose_reply /
confirm_reply. No env vars, no API key. Use this for interactive replies.
- Subprocess CLI (daemon path). For
reply bulk when no human is in the
loop. Auto-detects claude (Claude Code), cursor-agent, or codex on
PATH. Override with MEDIUM_OPS_LLM_CMD.
There is no paid-API-key path.
Auth setup
Medium has two auth layers that map to different feature surfaces:
- Integration Token —
Authorization: Bearer <token>. Used against
api.medium.com/v1/*. Gets you: publish_post,
list_own_publications. Token generation at
https://medium.com/me/settings → "Integration tokens".
Note: Medium stopped issuing new tokens in 2023. If you never
generated one, the write path will 401 and you'll have to use the
sid-cookie response path for any writes.
- sid cookie — from
medium.com (Application → Cookies → sid).
Used against medium.com/_/graphql and medium.com/_/api/*. Gets you:
all reads (stories, responses, claps, feed, stats, profile), plus
clap_post and post_response (fragile — undocumented).
medium-ops auth verify
medium-ops auth test
medium-ops auth setup
medium-ops auth har ./medium.har
Refreshing auth from a HAR
When cookies rotate or Medium changes a GraphQL schema, the fastest fix is:
- Open
medium.com in Chrome with devtools → Network panel.
- Reproduce the failing action (publish a draft, post a response, etc.).
- Right-click any request → "Save all as HAR with content".
medium-ops auth har ./medium.har
This:
- merges fresh
sid, uid, xsrf, cf_clearance cookies into .env
(preserving everything else)
- writes a redacted snapshot to
.cache/har-snapshot.json listing every
Medium GraphQL operation observed plus its request-variable / response-data
key shapes — useful for diffing against the queries hard-coded in
client.py to spot schema drift before users hit it.
Env vars (or ~/.cursor/mcp.json → mcpServers.medium-ops.env):
MEDIUM_INTEGRATION_TOKEN=2fb00...
MEDIUM_SID=1:...
MEDIUM_UID=...
MEDIUM_USERNAME=yourhandle
Architecture
mcp.json | env → auth.py
│
MediumConfig (token? sid? uid? username?)
│
MediumClient (httpx)
┌───────┼──────────┐
▼ ▼ ▼
api.medium.com medium.com/ medium.com/_/
/v1/* (REST) _/graphql api/* (dashboard)
│ │ │
Bearer token sid cookie sid cookie
│
┌──────┬──────┬────────────┬──────┬────────┬──────────┐
▼ ▼ ▼ ▼ ▼ ▼ ▼
posts responses claps profile stats feed reply_engine
│
┌───────────────────┼───────────────┐
▼ ▼ ▼
template ai_bulk MCP propose/confirm
└───────────────────┬───────────────┘
▼
base.post_response
│
┌─────────┼─────────┐
▼ ▼ ▼
dedup audit dry_run
(SQLite) (jsonl)
mcp/server.py ──── 22 tools ─── all share MediumClient
Endpoints used
| Auth: integration token | GET https://api.medium.com/v1/me |
| Auth: sid cookie | POST https://medium.com/_/graphql (Viewer) |
| User profile | POST /_/graphql (UserProfileQuery) |
| List stories | POST /_/graphql (UserStreamOverview) |
| Story metadata | POST /_/graphql (PostViewer) |
| Story body | POST /_/graphql (PostContent) |
| Story search | POST /_/graphql (SearchPosts) |
| Responses | POST /_/graphql (PostResponses) |
| Feed | POST /_/graphql (HomeFeed / FollowingFeed / TagFeed) |
| Publish story | POST https://api.medium.com/v1/users/{id}/posts |
| Publish to pub | POST https://api.medium.com/v1/publications/{pub_id}/posts |
| Own pubs | GET https://api.medium.com/v1/users/{id}/publications |
| Clap | POST https://medium.com/_/api/posts/{id}/clap (undocumented) |
| Post response | POST https://medium.com/_/api/posts (undocumented) |
| Stats | GET https://medium.com/@{username}/stats?count=... (undocumented) |
Tests
uv run pytest -q
Coverage: auth loading, client transports + XSSI stripping, dedup DB, audit
log search, MCP tool registry + dispatcher, MCP install host-config merging,
propose/confirm flow + token expiry, reply-engine template matching +
dedup+audit flow, subprocess LLM detection.
Known gaps
- Medium stopped issuing new Integration Tokens in 2023. If you never
got one,
publish_post / list_own_publications will 401. The
read + response + clap paths still work via sid. The RSS read path needs
no credentials at all.
- RSS is capped at ~10 posts and lacks clap/response counts and stats.
When you need more, pass
--source graphql (requires sid) or set
--limit > 10 and the client will auto-fall back to GraphQL.
- GraphQL operation names and schemas change silently. The queries in
client.py mirror what the dashboard uses today — expect breakage every
couple of months. Pin this package's version.
post_response uses GraphQL savePostResponse(deltas: [Delta!]!, inResponseToPostId: ID!). Delta shape is
{type: 1, index: N, paragraph: {type: 1, text, markups: []}} (type=1 means
insert; paragraph.type=1 is P). Reverse-engineered from error messages.
update_draft_content uses dashboard POST /p/{id}/deltas with
{baseRev, rev, deltas}. For a brand-new draft, baseRev=-1, rev=0.
Subsequent edits should bump both.
clap_post still uses the undocumented /_/api/posts/{id}/clap shape;
not yet re-validated against the new GraphQL surface. Dry-run first.
- Members-only stories return a paywall preview unless the
sid belongs
to a paying member.
- No "restack" equivalent. Medium doesn't have reshares; the closest is
a clap + a response. Use
clap_post + post_response together for that.
- No notes / short-form. Medium killed short-form in 2018.
- Chrome cookie auto-grab (the
auth_chrome flow from substack-ops) is
not yet implemented. Paste your sid into .env for now.
- TUI not yet implemented; the extras pin is there for future work.
License
MIT. See LICENSE.