Skip to main content

Support & Troubleshooting Assistant

User story​

As a Support Engineer working across large or distributed codebases, I want my AI assistant to quickly orient itself in unfamiliar repos and locate relevant code, configs, and docs — and when available, pull CI logs directly — so that I can troubleshoot issues without memorising every repo's layout or switching between tools.

Goals​

  • Agent orients itself immediately in any repo
  • Conventions surface the right runbook or doc for a given file path
  • Agent can search across multiple indexed repos and external sources
  • When CI logs are available, agent can fetch and parse them without leaving the chat

Technical overview​

Deployment: stdio (local). The agent already has workspace access via the host editor. mcpsmithy adds orientation and cross-repo search on top — it doesn't replace the agent's existing file tools.

Orientation: project_info gives the agent an immediate overview of the repo layout and what each source contains — so it can orient itself in an unfamiliar codebase without reading every directory.

Data sources:

  • Local source code with index: false — provides structure, not search
  • Remote runbooks and docs via project.sources.git — or project.sources.http for very large repos (pull a GitHub/GitLab archive instead of cloning)
  • External wikis or status pages via project.sources.scrape

Conventions: Map service paths to their runbooks so the agent knows what to read before investigating a given area.

Live API reads (optional): If your CI/CD system or any other service exposes an API, a tool using http_get lets the agent fetch data directly — CI logs for a given pipeline, Slack channel history for a support thread, or a status page summary. Use grep in the template to filter the output down to relevant lines before it reaches the agent.

Tools needed:

  • project_info — repo layout and source descriptions
  • find_convention — returns the docs, convention rules, and workflows that apply to a file path
  • search — ranked search across all indexed sources; conventions are surfaced first so the agent gets the right context before digging into code
  • ci_log — fetches CI/CD logs for a given run or pipeline via http_get
  • slack_history — fetches recent messages from a Slack channel via http_get; useful for reviewing support threads or incident chatter

For full YAML examples of each source type, tool template, and convention pattern, see the config reference.

Generate this config with your agent

Run mcpsmithy setup, then share this story with your agent. Before prompting, have ready:

  • A list of the repos to index (runbooks, service docs, or both)
  • Which service paths map to which runbook or doc area
  • If you have a CI API: the endpoint and any auth needed for log fetching

Then use a prompt like:

Set up mcpsmithy for support troubleshooting across this platform. Here are the repos to index: [list them]. Map each service path to its runbook via conventions. Expose project_info, find_convention, search, and ci_log tools (ci_log is optional — only if I provide a CI API endpoint).

See Assisted setup for the full workflow.

Adapting this for user-facing support

With minimal changes this config works for end-user support too — even when you have no dedicated docs site. The main difference is which files you index: narrow the source globs to docs, READMEs, and changelogs (exclude internal implementation files), and add a read_doc tool so the agent can surface specific files directly. The repo becomes your support knowledge base with no extra tooling.