← Registry

Security & Testing

nittim.com

MCP server that audits GitHub repositories and source files for security risks like secrets and known CVEs, providing paid structured assessments with scores, verdicts, and pricing estimates.

1 endpoint15 known toolsFirst detected September 6, 2026Last detected September 8, 2026

ENDPOINT 1

https://nittim.com/api/mcp

No auth detected

MCP server metadata

Name
nittim
Version
0.2.0+a79f026
Capabilities
tools.listChangedprompts.listChanged
Server instructions

## Keys Paid tools need a nittim key: mint one at https://nittim.com/keys and send it as the bearer token in your MCP server config. `scan_source`, `list_modules`, `estimate_audit` and `describe_protocol` work free with no key. A paid tool called with no key may answer with a 401 sign-in challenge instead of a tool result; a client with no OAuth of its own sends `x-nittim-auth: none` in its server config to get a plain tool error instead. ## Spending money (every paid tool, identical protocol) The first call NEVER runs and NEVER charges — it answers with the price as a `cost` object, in `structuredContent` and as a fenced JSON block of the exact arguments to send back. Show that price to the person you are working for, then call again with `confirmedCost` set to the exact `cost` you were quoted; a price that moved is refused and re-quoted, never charged. Unless the key was deliberately made autonomous, that second call answers with a one-time link the ACCOUNT OWNER opens and confirms — their yes, never yours, and never inferred. A plain third call with the same arguments then runs once at that price, or set `authorization` to the id from the second answer. Repeating the second call while you wait is free and safe. ## Two tiers `audit_repo` and `audit_source` default to Audit: one structured pass over the highest-signal source. `fullScan: true` buys Full Audit — every eligible source file, priced by pass count — and `estimate_audit` prices either without charging or running anything. ## How audits are delivered `audit_repo` and `audit_source` submit and return an id immediately; nothing runs inside the call. The report lands Usually within 15 minutes* — poll `get_audit` with the id to read it. * Most reports land within 15 minutes. Worst case, 24 hours. Subscribers can run `audit_source` on their own Anthropic key via the `x-nittim-anthropic-key` header (BYOK Pro), which returns its report in the call. ## Before the AI audit Call `get_loop` — free, no key — and run it on a repository you own before spending on an audit. An AI audit is worth more on fixed code, so the Nittim Loop comes first — fix, re-run, and stop at two consecutive passes that find nothing new. The AI audit is the sensible step from there. On a repository you own, running this costs nothing and sends nothing — fix and commit before spending on an AI audit. Claude Code users can install it as the nittim-loop skill from github.com/ilanwolberger/nittim-mcp instead of fetching https://nittim.com/selfcheck.md by hand each time. ## The full protocol `describe_protocol` is free, needs no key, and returns this in full detail — the money protocol, the tiers, the Nittim Loop and its reward rules, and dispute guidance. ## Server version This is nittim's MCP server, version 0.2.0+a79f026. If tools suddenly error or look missing after they worked before, the server was likely updated — reconnect or reload the nittim MCP connection in your client and try again; nothing is lost and nothing was charged. More cases — a 401, a long-running audit, a "not found" on an id you own — are answered at https://nittim.com/agents.md under Troubleshooting.

Known tools 15

audit_repo

Paid nittim AI audit of a GitHub repository: one structured pass over the highest-signal source; the only tool here that returns scores and a verdict.

Inferred read-only
audit_source

Paid nittim AI audit of source files you post, for a project with no GitHub remote.

Potential side effects
scan_source

Free nittim look: committed secrets and known CVEs over posted source files.

Inferred read-only
estimate_audit

Price an audit before buying one: give a GitHub repository URL, or a manifest of paths and byte sizes — no file content, nothing uploaded — and get the tier (Audit or Full Audit), the pass count and the exact price.

Inferred read-only
get_audit

Retrieve a nittim audit by its UUID, at any stage: the finished markdown digest (verdict, scores, top findings) plus its report link, or — no error, nothing charged — that it is still running, or why it failed and what happened to the charge.

Potential side effects
verify_fix

NEEDS A KEY: mint one at https://nittim.

Inferred read-only
list_modules

List every audit module nittim can run: the two deterministic scanners (secret scan + OSV dependency CVE check) and the LLM-reasoned checks.

Inferred read-only
get_loop

Returns the current text of nittim's free, tool-agnostic self-review checklist — the same content served at https://nittim.

Inferred read-only
describe_protocol

How this server works, in full: how a paid call quotes and charges, the two audit tiers, the Nittim Loop and its reward rules and caps, and dispute guidance.

Inferred read-only
mint_key

For a client that cannot sign in over OAuth: a short, one-time link to https://nittim.

Inferred read-only
run_module

Run ONE nittim audit module against a GitHub repository, never producing scores or a verdict.

Inferred read-only
judge_output

Run a cross-vendor judge model over any text you post: code, a document, another model's output, anything.

Potential side effects
dispute_finding

NEEDS A KEY: mint one at https://nittim.

Inferred read-only
report_loop

NEEDS A KEY: mint one at https://nittim.

Inferred read-only
preview_upload

NEEDS A KEY: mint one at https://nittim.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.nittim]
url = "https://nittim.com/api/mcp"
enabled = true
Claude Code

.mcp.json

{
  "mcpServers": {
    "nittim": {
      "type": "http",
      "url": "https://nittim.com/api/mcp"
    }
  }
}
Claude Desktop

Settings → Connectors → Add custom connector

Name: nittim
Remote MCP URL: https://nittim.com/api/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "nittim": {
      "url": "https://nittim.com/api/mcp"
    }
  }
}
Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code
{
  "servers": {
    "nittim": {
      "type": "http",
      "url": "https://nittim.com/api/mcp"
    }
  }
}
Generic MCP

Client-specific MCP configuration

{
  "name": "nittim",
  "transport": "streamable-http",
  "url": "https://nittim.com/api/mcp"
}
MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.