HackMD for AI agents
HackMD is a collaborative markdown workspace for writing and sharing personal and team notes and documents. Teams use it to handle this step of their document workflow without extra manual tooling. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in HackMD.
Create a note from content
Create a new markdown note from meeting minutes or drafted text for the team to review.
Read an existing note
Pull the content of an existing note to summarize it or reuse it elsewhere.
Restore a trashed note
Restore a note from trash only after confirming with its author that it was deleted by mistake.
8 HackMD actions, graded by risk.
Every HackMD action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
4Look things up. Allowed by default.
- hackmd.get_noteGet Note
- hackmd.list_notesList Notes
- hackmd.get_accountGet Account
- hackmd.list_foldersList Folders
write
2Create and change things. Allow, or ask the user first.
- hackmd.create_noteCreate Note
- hackmd.restore_notesRestore Notes
destructive
2Delete, cancel or archive. Ask first, or deny outright.
- hackmd.list_trashList Trash
- hackmd.move_note_to_trashMove Note to Trash
HackMD in three steps.
- 01Your users connect HackMDThey add their HackMD api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, writes like “create Note” can wait for the user, and “list Trash” can be denied outright.
- 03Any agent can actYour agent calls HackMD through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('hackmd', { read: 'allow', write: 'ask', // create_note destructive: 'deny', // list_trash }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How HackMD connects.
Users add their HackMD api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same HackMD connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use HackMD from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
HackMD and Arc0, answered.
Can I use HackMD with Claude, ChatGPT or Cursor?
Yes. Connect HackMD to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the HackMD actions you allow.
How do users connect HackMD?
Users add their HackMD api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which HackMD actions can my agent take?
8 in total: 4 read, 2 write and 2 destructive, such as “create Note”. Your policies decide which of them each agent may call.
Can I stop my agent from deleting things in HackMD?
Yes. Actions like “list Trash” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.
Can my own backend call HackMD too?
Yes. The same HackMD connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug HackMD into your agent.
Your users connect HackMD once, under your brand. Your agent gets 8 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan