← Developer toolsHosted MCP

A little context.
For your assistant.

Let a compatible AI client retrieve your assessments and work with the evidence gaps you’ve identified.

Connect your client

  1. Create a key in your account. Enable mcp and read for retrieval; add write only if you want to submit new transcripts.
  2. Use a client that supports remote MCP over Streamable HTTP with a Bearer authorization header.
  3. Add the endpoint and header below in that client’s connection settings. The exact settings vary by client.
Endpoint: https://badcop.tech/mcp/
Transport: Streamable HTTP
Authorization: Bearer YOUR_API_KEY

If your client only supports local processes or OAuth connections, this Bearer-token setup will need a compatible connector. Never put your key into a shared prompt or commit it to a repository.

Available tools

ToolUse it toArguments
list_sessionsFind your assessments.page, per_page (optional)
get_insightsRead prepared findings and summary.session_id
get_reportRead an unlocked, prepared report.session_id
analyze_transcriptCreate an assessment from interview text.transcript (at least 50 characters)

Try a useful question

Find my latest assessment and summarize the evidence I still need to gather. Distinguish what I reported from what remains unverified.

Before sharing a report

Your client and its AI provider may process any content retrieved through these tools. Check their data settings and share only information you are authorized to provide.

Analysis can take a few minutes and creates a new assessment each time. If a request times out, check your saved assessments before retrying. Report access and daily key limits also apply here.

MCP provides access to your badcop data. It does not independently inspect your code, validate your infrastructure or certify your technology.
API permissions and errors →