Mcp Server — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Mcp Server (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Bring Summit53's Revenue Engine into the AI workflows your team already uses.
This package connects MCP-compatible assistants to Summit53's hosted Revenue Engine. Assistants can help diagnose pipeline and forecast gaps, decide the next best action, execute permitted workflows, and preserve value evidence around the Figure-Eight revenue loop — from acquisition through expansion and back into acquisition.
Using Claude or Codex? Install the Summit53 plugin from github.com/Summit53/skills instead. It connects the hosted Summit53 server directly (no local package) and adds guided Summit53 skills for forecast, deal, renewal, coaching, territory and board workflows.
>
```text # Claude Code /plugin marketplace add Summit53/skills /plugin install summit53@summit53
>
# Codex codex plugin marketplace add Summit53/skills codex plugin add summit53@summit53 ```
>
In claude.ai, Claude Desktop or Cowork, add the marketplace https://github.com/Summit53/skills under Customize › Plugins › Add. This repository is not a plugin marketplace.Most clients, including claude.ai, Claude Desktop, Claude Code, Codex and Cursor, can connect to the hosted server directly at https://api.summit53.io/mcp and sign in with OAuth:
claude mcp add --transport http summit53 https://api.summit53.io/mcp
codex mcp add summit53 --url https://api.summit53.io/mcpThis package is a local stdio bridge for clients that only support command-based servers:
{
"mcpServers": {
"summit53": {
"command": "npx",
"args": ["-y", "@summit53/mcp-server"]
}
}
}On first use, your browser opens the Summit53 OAuth flow. Sign in, review the requested access, and authorize the client. Tokens are cached by the local MCP bridge and refreshed automatically.
For a web client that supports custom remote MCP connections, add:
Summit53https://api.summit53.io/mcpClient availability and labels vary by plan and workspace. Use the custom server URL instead of relying on a public app-directory listing.
See the MCP setup guide for client-specific instructions.
Ask in natural language to:
The current release gate freezes 75 externally published functions across 15 categories. That count is a compatibility fact, not the product proposition. Tool visibility depends on the client's granted scopes and the user's Summit53 organization permissions; call list_available_tools to discover the surface available in the current session.
| Category | Published functions |
|---|---|
| Search | 3 |
| Opportunity Strategy | 5 |
| Account Intelligence | 3 |
| Pipeline & Risk | 4 |
| RevOps | 2 |
| Forecasting | 9 |
| Activity | 3 |
| Research | 1 |
| Notes & Activities | 5 |
| Outreach | 17 |
| Value Impact | 11 |
| Coaching | 3 |
| Knowledge Base | 5 |
| Job Lifecycle | 3 |
| Discovery | 1 |
The exact machine-readable surface is maintained in tool-catalog.json. The registry-compatible connection manifest is server.json. The tool reference explains inputs, outputs, access requirements, and async behavior.
Interactive connections use OAuth 2.1-style authorization with PKCE. Access tokens last one hour; rotating refresh tokens last seven days. Access can be revoked from /admin/integrations/mcp in Summit53.
The published scope set is:
crm:read, pipeline:read, revenue:read, actions:read, jobs:read;crm:write, outreach:write, outreach:send, coaching:write, value:write, automation:manage, jobs:manage.Business write tools always require their explicit write scopes. Job lifecycle tools always enforce jobs:read or jobs:manage. Read-scope enforcement for the remaining catalogue is being rolled out behind a production gate. Summit53's organization, ownership, and role permissions remain authoritative in every case.
Some approved service integrations can use scoped static API keys, but API keys are not part of the normal interactive setup. A scope authorizes an operation; it does not itself prove human approval. Broader high-risk write workflows remain gated until portable approval controls are available.
Long-running tools return a durable job response rather than asking the client to hold an HTTP request open. Use:
get_job to inspect status and results;list_jobs to find recent work;cancel_job to request cancellation when the job is still cancellable.Job records are retained for 30 days. Every published function uses the same versioned response envelope so clients can handle success, partial results, queued work, and errors consistently.
https://api.summit53.io/mcpSummit53 is working with a small group of B2B SaaS revenue teams to shape the Revenue Engine in production. Apply to the Founding 50.
Proprietary. See summit53.io/terms-of-service for details.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.