mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-10-21 00:05:49 +02:00
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - People also use assistants in Codex, Claude, and other MCP clients. > - The existing assistant connection can read work and create tasks or comments. > - It cannot edit tasks, exchange files, or manage normal agent and project settings. > - These operations must retain the person's permissions and Paperclip's execution rules. > - This pull request adds an explicit operation registry and separately consented configuration access. > - Assistants can manage work without receiving credentials or runner authority. ## Linked Issues or Issue Description **Subsystem affected** Cross-cutting: assistant MCP, domain routes, consent UI, storage, and Product E2E. **Problem or motivation** A connected assistant cannot update tasks, maintain documents, attach files, or configure existing agents, projects, and skills. Users must leave the assistant for these routine actions. **Proposed solution** Add named tools and a restricted API registry to direct connections. Require separate configuration consent. Reuse domain routes and retry receipts. File uploads save the attachment when the byte transfer succeeds. **Alternatives considered** Arbitrary REST forwarding would expose administration and credential operations. Runner impersonation would bypass execution ownership. Separate upload completion calls add unnecessary client state. **Roadmap alignment** Checked ROADMAP.md and related MCP pull requests. This extends the human-authorized connection from #14933. It does not replace the runner or introduce agent impersonation. Companion Cloud routing and directory isolation: https://github.com/paperclipai/paperclip-cloud/pull/678. ## What Changed - Add task editing, finish/block, documents/revisions, deliverables, agent settings/instructions, projects/repositories, and skills/files. - Add an allowlisted API search/call registry with identical field restrictions, scopes, and retry identities. - Add unchecked configuration consent. Existing write grants retain their current authority. - Add hashed, expiring file transfer tickets and atomic upload receipts. No completion call is required. - Preserve company boundaries, human attribution, active native execution ownership, and execution review gates. - Add protocol/domain tests, consent stories, and eight paid Product E2E workflows. - Repair two CI fixture races: await cold route setup before assertions, and wait for asynchronously loaded connection copy. Both fixture suites pass (24 + 48 tests). ## Verification - Consent revision: one write-access checkbox controls requested work and configuration permissions in browser and device flows. All 16 consent tests, UI typecheck/build and token gates pass. Updated interactive stories cover default approval, opt-out and viewer restrictions. The paid browser helper uses the new exact label. Real GPT-5.4 Mini Product E2E passes 2/2 at `64f96373118eb190f8cba1c2ab17cb979555f3ad` (configuration + permission denial), campaign `local-2026-10-07T00-51-14-337Z`, no automatic retries, cleanup passed; $0.04149375 estimated assistant cost plus unpriced worker usage. Raw results, usage and source fingerprints are retained in the worktree. UI and Product E2E typechecks pass. - Prior head `2f246d4b74f1f98c75ebcb37ae6753a748237fac`: all 52 checks pass; two optional Storybook checks skip. Greptile 5/5 on that head, no unresolved review threads. Final consent head `64f96373118eb190f8cba1c2ab17cb979555f3ad` also has all checks passing and Greptile 5/5 with no unresolved threads. The unchanged Cursor sandbox test had one 10-second timeout, passed in local isolation, and passed its single CI rerun; the failed attempt remains in [the CI run](https://github.com/paperclipai/paperclip/actions/runs/37554106934). The existing chat retry-denial browser test had one visibility failure; its single rerun passes, and the failed attempt remains in [the CI run](https://github.com/paperclipai/paperclip/actions/runs/37542735691). - Full workspace `pnpm -r typecheck` and `pnpm build` pass at final runtime source `b2196fae1`. UI token gates pass. - 139 MCP/OAuth/transfer/privacy tests and 76 grader calibration tests pass, including one-connection PostgreSQL OAuth and concurrent upload retries. - Paid Product E2E: all eight expanded cases qualified across Mini, Haiku and Sonnet. A merged-source repeat passed 23/24; one Haiku cell timed out before application startup. Final affected-case qualification passes 9/9 on all three models with grader v16, including the failed cell. Automatic retries disabled; failures, costs, source hashes and independent durable-state/file assertions are retained in [the verification record](doc/plans/2026-10-06-expanded-assistant-mcp-verification.md). - Actual Codex CLI, Claude Code and OpenCode clients completed local reads/mutations. Codex wrote a report, Claude updated it in a later conversation, and OpenCode uploaded/downloaded a file with matching SHA-256 and registered the attachment. Revoking the CLI grant rejects subsequent bridge initialization. - Butter staging is verified on final runtime `b2196fae1` ([deployment](https://github.com/paperclipai/paperclip-cloud/actions/runs/37538432138)). A fresh OpenCode workspace fetched the copied invitation, configured remote MCP, started OAuth and reached real consent with configuration unchecked. Invalid transfer tickets return 403 through Cloud. Human approval for the new persistent staging grant is pending; hosted task/file success is not yet claimed. The final transaction fix is deployed. - Full local `pnpm test:run` passed 15,614 general-server tests but stopped on two macOS timeouts. The heartbeat test passed in isolation; the existing 40,000-file Git stress fixture timed out again. Its Linux CI lane passes. Later local full-suite phases did not run after the timeout; this is not an all-green local full-suite claim. - Instructions and security limits are in `doc/public-mcp.md`; the saved plan is `doc/plans/2026-10-06-expanded-assistant-mcp-tools.md`. ## Risks - This expands the experimental direct MCP surface. Explicit schemas and domain permissions must stay synchronized. - Migration 0311 adds transfer tickets and upload receipts. Expired orphan cleanup must not remove committed attachments. - Configuration requires a new consent request containing that scope; the single write-access choice controls it alongside work mutations. Refreshing an old grant does not add it. - The public directory keeps its original ten tools through the companion Cloud change. - Hosted consent/work proof remains the final delivery gate. The PR stays draft while approval of the new staging grant is pending; code checks and review are green. Merging is a separate action. ## Model Used OpenAI Codex (GPT-6, tool use and code execution). The exact serving model ID and context window are not exposed in this session. Paid evaluation models: gpt-5.4-mini, claude-haiku-4-5-20251001; claude-sonnet-4-6. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass (focused suites; full-suite macOS limitation disclosed above) - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
63 lines
6.8 KiB
TypeScript
63 lines
6.8 KiB
TypeScript
import { PUBLIC_MCP_SCOPES } from "./public-mcp.js";
|
|
|
|
export type AssistantClient = "codex" | "claude" | "opencode" | "browser" | "headless" | "other";
|
|
export const assistantClientNames: Record<AssistantClient, string> = {
|
|
codex: "Codex", claude: "Claude Code", opencode: "OpenCode", browser: "ChatGPT / Claude web", headless: "Headless", other: "Other",
|
|
};
|
|
export function mcpSetupUrl(serverUrl: string, companyId?: string) {
|
|
const url = new URL("/mcp/setup", serverUrl);
|
|
if (companyId) url.searchParams.set("company", companyId);
|
|
return url.toString();
|
|
}
|
|
export function mcpInvitation(serverUrl: string, company?: { id: string; name: string }) {
|
|
return `Connect to my ${company ? `${company.name} ` : ""}Paperclip organization using the instructions at ${mcpSetupUrl(serverUrl, company?.id)}. Start authorization, give me the approval link, and verify the connection after I approve. This connects as me, not as a Paperclip agent.`;
|
|
}
|
|
export function mcpSetupSteps(serverUrl: string, assistant: AssistantClient) {
|
|
const quoted = `'${serverUrl.replaceAll("'", "'\"'\"'")}'`;
|
|
const scopes = PUBLIC_MCP_SCOPES.join(" ");
|
|
switch (assistant) {
|
|
case "codex": return [
|
|
{ text: "Check for an existing Paperclip server first. Reuse it only if its URL matches this instance. Add the remote server with the Codex CLI.", code: `codex mcp add paperclip --url ${quoted}` },
|
|
{ text: "Start browser authorization and give the user the approval link.", code: "codex mcp login paperclip --scopes paperclip:read,paperclip:write,paperclip:configure,offline_access" },
|
|
{ text: "If the current conversation does not reload its MCP tools, start a new conversation or restart the client. Verify the connection with paperclip_connection before claiming success." },
|
|
];
|
|
case "claude": return [
|
|
{ text: "Check existing servers and reuse a matching instance. Add the remote server using Claude Code.", code: `claude mcp add --transport http paperclip ${quoted}` },
|
|
{ text: "In Claude Code, run /mcp, select paperclip, and choose Authenticate. Versions that offer `claude mcp login --help` can also start authorization with `claude mcp login paperclip`. Give the user the approval link." },
|
|
{ text: "Reconnect through /mcp, or start a new session if needed. Verify the account and organization with paperclip_connection." },
|
|
];
|
|
case "opencode": return [
|
|
{ text: "Check `opencode --version` and existing configuration. For OpenCode 1.x, merge this entry into opencode.json without replacing other settings. OpenCode 2.x uses mcp.servers; use that version's documented format instead.", code: JSON.stringify({ mcp: { paperclip: { type: "remote", url: serverUrl, enabled: true, oauth: { scope: scopes } } } }, null, 2) },
|
|
{ text: "Run from the same project directory. OpenCode opens Paperclip's sign-in and consent page; give the user the approval link.", code: "opencode mcp auth paperclip" },
|
|
{ text: "If OpenCode was already running, restart it to load the connection. For the browser app, use this command. Verify the tools with paperclip_connection.", code: "opencode web" },
|
|
];
|
|
case "browser": return [
|
|
{ text: "Open your assistant's Apps / Connectors settings and add a custom remote MCP connection using this URL. Availability depends on the account and workspace policy. Reading this page in chat does not install a connector.", code: serverUrl },
|
|
{ text: "Choose OAuth / browser sign-in. Approve the connection in Paperclip, then select the connector in your conversation. If custom connectors are unavailable, use a supported local client. Verify the connected account and organization before doing work." },
|
|
];
|
|
case "headless": return [
|
|
{ text: "Use a Paperclip CLI version with `mcp login --device` support. This needs no callback listener; it prints a verification URL and user code, then waits for human approval.", code: `paperclipai mcp login --device --url ${quoted}` },
|
|
{ text: "After approval, configure a local stdio MCP server running the command below. Credentials stay in the CLI's private credential store. Do not copy tokens or device codes into chat or project configuration.", code: `paperclipai mcp proxy --url ${quoted}` },
|
|
{ text: "The user code is for the human to verify the request. The private device code remains inside the CLI. After connecting, call paperclip_connection to verify the account and organization." },
|
|
];
|
|
default: return [
|
|
{ text: "Add a remote HTTP MCP server at this URL, then choose OAuth / browser sign-in.", code: serverUrl },
|
|
{ text: "If the client cannot install MCP servers from chat, use its connection settings. If it cannot receive a browser callback, use the Headless instructions. After approval, verify the account and organization with paperclip_connection." },
|
|
];
|
|
}
|
|
}
|
|
|
|
export const mcpAuthorizationHandoffInstructions = "Authorization commands wait for human approval. When using an assistant shell tool, keep the command running in a persistent terminal or background process, capture its output privately, and share the approval URL immediately while it is still running. Client callback deadlines still apply. If the command timed out or stopped, start a fresh authorization request before sharing a link. If the host cannot keep the command alive across turns, ask the user to run it in their own terminal.";
|
|
|
|
export function mcpSetupMarkdown(serverUrl: string, companyId?: string) {
|
|
return ["# Connect your assistant to Paperclip", "",
|
|
"These are setup instructions, not an access credential. Ask the human to approve the identified client in Paperclip. Never approve on their behalf or reuse browser cookies as MCP credentials.", "",
|
|
`MCP server: ${serverUrl}`, ...(companyId ? [`Requested organization ID: ${companyId} (a hint, not authority).`] : []), "",
|
|
"Keep existing client settings. If the name paperclip already points elsewhere, use a distinct server name and the same name in the login command. Stop and explain if the host cannot configure MCP. A restart or new conversation may be required; do not claim tools are available until a real call succeeds.", "",
|
|
mcpAuthorizationHandoffInstructions, "",
|
|
...Object.entries(assistantClientNames).flatMap(([client, name]) => [`## ${name}`, "", ...mcpSetupSteps(serverUrl, client as AssistantClient).flatMap((step) => [step.text, ...(step.code ? ["", "```", step.code, "```"] : []), ""])]),
|
|
"## After approval", "",
|
|
"Call paperclip_connection, verify the human and organization, then summarize agents and open tasks. If the approved organization differs from the requested one, stop and reconnect. Do not create tasks or comments merely to test a connection. Creating tasks or comments can start paid agent work; do that only when requested. Access is governed by the consenting human's permissions, not a selected agent's identity.", "",
|
|
].join("\n");
|
|
}
|