10gen Repo Intelligence
10gen Repos — Case Tracker and Troubleshooting Extension Context
When to use this skill
- Identify which 10gen repo or script is most relevant to a case-tracker or troubleshooting scenario
- Get install/run guidance for any repo in the high-signal set
- Find integration ideas for connecting external repos into the Chrome extension or local backend
- Map a scenario (FTDC diagnostics, explain plans, Jira enrichment, escalation UI) to the right tool
- Understand which repos require internal credentials or special prerequisites
When NOT to use this skill
- General Atlas troubleshooting with no 10gen repo angle — use
atlas-diagnostics-expert - Case management and tracker workflows — use
case-tracker - Extension diagnostic UI patterns — use
atlas-diagnostics-expert
Skill guidance
- Use the scenario-to-repo mapping section as the first lookup for any incoming question.
- Cross-reference
docs/init_index.mdfor broad repo discovery anddocs/high_signal_file_index.jsonfor per-file detail. - Prefer backend-mediated integrations over direct browser-side secrets.
Highest-value repos
mdb-tam
Direct Chrome extension architecture, local backend, native messaging hosts, Monday sync, Glean/Slack/calendar integrations, operator dashboards.
- Best scenario: Build or extend the extension, mirror customer context locally, or automate operator workflows.
- Install:
npm install; npm run prepare:extension-root; cd native-host && ./install.sh <extension-id> - Run: Load repo root as unpacked Chrome extension; optional Node CLIs via
node scripts/*.mjs - Notes: Strongest reference implementation. Many integrations should be backend-mediated rather than direct browser calls.
ts-tools-customer-hub
Customer context UX, Salesforce/Coveo/S3 integrations, monitoring, and runbook patterns.
- Best scenario: Account context views, search, escalation workflows, and complex operator navigation patterns.
- Install:
npm ci --ignore-scriptsafter private npm auth and env setup - Run:
npm run dev - Notes: Large, high-value repo; reuse concepts selectively.
ts-tools-support-api
Case creation, uploads, auth flows, swagger/openapi-backed automation surface.
- Best scenario: Secure backend for programmatic case operations.
- Install:
npm install - Run:
npm run devafter configuring Salesforce/JWT-related env vars - Notes: Security-sensitive. Prefer a thin backend wrapper instead of exposing it directly to the extension.
ts-tools-scrapers-tickets
Jira, Salesforce, and support-alert ingestion; support DB synchronization patterns.
- Best scenario: Background enrichment, queue hydration, and joining ticket streams into one support model.
- Install:
poetry install - Notes: Good ingestion logic; fits better behind a backend worker due to scheduling and service credentials.
ts-tools-escalation-manager
Escalation-oriented UI, list/detail flows, active escalated case handling.
- Best scenario: Extension side-panel UX around escalations or “cases needing attention now.”
- Install:
npm install --legacy-peer-deps - Run:
npm run dev - Notes: UI-focused; easier to mine for patterns than ts-tools-customer-hub.
ts-search-explain-helper (SearchPlanIQ)
Atlas Search explain diagnostics, visual exports, markdown writeups, MCP server mode.
- Best scenario: Cases involving Atlas Search or vector search that need explain-plan analysis.
- Install:
uv syncoruv pip install -e . - Run:
uv run python -m searchplaniq.apporsearchplaniq-mcp - Notes: High-value troubleshooting tool; great fit for attachment-based workflows and AI summarization.
alexandria
FTDC diagnostics, rule-based issue detection, metric explorer, remote web analysis.
- Best scenario: FTDC-driven troubleshooting and automated “what stands out in this diagnostic bundle?” flows.
- Install:
go build ./...or use a release binary - Run:
./alexandria <diagnostic.data path> - Notes: Best FTDC-specific diagnostic engine in the set.
devprod-mcp-router
Unified access to Jira, Confluence, Evergreen, Git, Backstage, and other debugging tools through one MCP gateway.
- Best scenario: Extension/backend needing broad tool access without bespoke integrations.
- Install:
go build ./...orgo installthedevprod-mcp-proxycommand - Run:
devprod-mcp-proxy --gateway <gateway-url> - Notes: Ideal as a force multiplier for agentic flows.
ts-support-demand-tracking
Support demand signals from Trino, Gmail, Discord, and MongoDB aggregation.
- Best scenario: Prioritization, trend detection, hot-topic identification, or staffing/context overlays.
- Install:
uv sync - Run:
python support_demand_tracker.py
mongolyser
Local desktop analysis of logs, profiling, indexes, cache, sharding, and cluster health.
- Best scenario: Companion deep-dive tool when extension triage decides a human needs full diagnostics.
- Install:
npm install - Run:
npm run electron:start - Notes: Better as a companion tool than something embedded in-browser.
hmm-aha-search
Aha ticket and feature search with hybrid retrieval and assistant endpoints.
- Best scenario: Roadmap/product context and feature history during customer issue work.
- Install:
bun install - Run:
bun dev
ts-ftdc-requestor
Requesting FTDC from Atlas and storing/retrieving shared FTDC bundles.
- Best scenario: “Request diagnostics now” or collaboration on FTDC bundles.
- Install:
pip install -r requirements.txtor Docker-based local setup - Notes: Strong backend action surface for extension-triggered diagnostics collection.
Useful later / medium-signal repos
| Repo | Purpose |
|---|---|
support-dashboard |
Legacy dashboard patterns for hot-customer and queue views |
ts-tools-diagnostic-viewer |
Diagnostic artifact viewer UI patterns |
ts-tools-diagnostics-pipeline |
Async diagnostics processing / contributor APIs |
support-tools |
Large hunting ground for niche internal support utilities |
scripts-and-snippets |
Internal script catalog and discovery point |
atlas-tools |
Atlas node-side support scripts |
ts-query-helper |
Older explain-analysis app and deploy scripts |
ts-tools-calendar |
Calendar automation microservice |
ts-tools-workflow |
Workflow engine with approvals and queues |
ts-tools-hub |
Legacy support hub triage UI |
agent-skills |
Procedural knowledge/skill pack for future AI assistance |
ai-tools-code-samples |
LLM and MongoGPT integration examples |
Scenario-to-repo mapping
| Scenario | Primary repos | Start with |
|---|---|---|
| Create or update a support case | ts-tools-support-api, mdb-tam, ts-tools-escalation-manager | ts-tools-support-api automation endpoints |
| Enrich a case with Jira/Salesforce/support-alert context | ts-tools-scrapers-tickets, ts-tools-customer-hub, devprod-mcp-router | Jira/SFDC scrapers |
| Collect or analyze FTDC diagnostics | ts-ftdc-requestor, alexandria, ts-tools-diagnostic-viewer | FTDC request API; then Alexandria CLI |
| Analyze Atlas Search or vector search explain plans | ts-search-explain-helper, ts-query-helper | SearchPlanIQ app and MCP server |
| Prioritize cases by support load or signal volume | ts-support-demand-tracking, support-dashboard | Support volume tracker |
| Retrieve runbooks, knowledge, or engineering context | devprod-mcp-router, hmm-aha-search, agent-skills | Confluence/Git/Backstage via MCP |
| Schedule follow-up actions | ts-tools-calendar, mdb-tam | CalGuru calendar API |
Top scripts / tools
| Script / tool | Repo | Purpose |
|---|---|---|
| Customer Dashboard extension | mdb-tam | Direct blueprint for the target product |
server/src/index.js |
mdb-tam | Local backend entrypoint for SSE, reports, and integration mediation |
native-host/install.sh |
mdb-tam | Native messaging host installer |
scripts/monday-initiative-sync.mjs |
mdb-tam | Monday sync CLI |
scripts/getPickListValues.ts |
ts-tools-customer-hub | Salesforce picklist sync utility |
| Support API endpoints | ts-tools-support-api | Programmatic case automation and auth surface |
jira_scraper.py |
ts-tools-scrapers-tickets | Jira ingestion for support DB synchronization |
support_demand_tracker.py |
ts-support-demand-tracking | Aggregates support demand signals |
searchplaniq CLI |
ts-search-explain-helper | Explain analyzer entrypoint |
searchplaniq-mcp |
ts-search-explain-helper | MCP server for explain-plan tooling |
alexandria CLI |
alexandria | FTDC diagnostics and rule evaluation |
devprod-mcp-proxy |
devprod-mcp-router | Client-facing gateway proxy for MCP backends |
npm run electron:start |
mongolyser | Desktop diagnostic workstation start command |
Recurring mechanisms and APIs
| Mechanism | Key repos | Notes |
|---|---|---|
| Chrome Extension / MV3 | mdb-tam | Key for browser-native operator workflows and side panels |
| Native Messaging | mdb-tam | Lets a Chrome extension talk to local Python/shell hosts |
| Salesforce | ts-tools-support-api, ts-tools-scrapers-tickets, ts-tools-customer-hub | Source of support case data, auth flows, and customer metadata |
| Jira | ts-tools-scrapers-tickets, devprod-mcp-router | Issue and escalation context |
| Atlas API | ts-ftdc-requestor, alexandria | Used for FTDC requests and cluster diagnostics |
| FTDC | alexandria, ts-ftdc-requestor | Primary diagnostic artifact for storage/performance troubleshooting |
| Explain plans / Atlas Search | ts-search-explain-helper | Central to search troubleshooting |
| MCP | devprod-mcp-router, ts-search-explain-helper | Tool aggregation and agent bridge pattern |
Implementation guidance
- Prefer backend-mediated integrations for Salesforce, Jira, Slack, Atlas, Gmail, Discord, and other privileged systems. Avoid direct browser-side secrets.
- Treat
mdb-tamas the primary architectural reference for extension composition, local/backend mediation, prompt workflows, and operator UX. - For diagnostics, keep the extension lightweight: trigger actions through
ts-ftdc-requestor,alexandria,ts-search-explain-helper, ormongolyserinstead of rebuilding those engines in-browser. - Use
devprod-mcp-routerwhen broad engineering-tool access is needed. - Mine
ts-tools-customer-hub,ts-tools-escalation-manager, andsupport-dashboardfor workflow ideas and UI patterns; avoid large-scale code copying unless dependencies are already aligned.
Caveats and risk notes
- Many repos require internal credentials, Kanopy/CorpSecure access, private npm auth, Atlas API keys, Salesforce integration users, or Google service accounts.
- Some tools are legacy, Stitch/App Services-era, or partially documented. Validate fit before adopting them as core dependencies.
- Diagnostic artifacts (FTDC, explain plans, logs, support tickets) may contain sensitive customer or operational information. Keep storage, transmission, and sharing paths explicit.
- Not every useful repo should be embedded directly. Some are better as external companion tools or backend services invoked by the extension.
Installation and setup
The generated script install_10gen_case_tracker_dependencies.sh is clone-first and safe by default:
- Clones high-signal repos into
~/Documents/GitHub/10genand installs prerequisite toolchains only when missing and when Homebrew is available. - Dependency installation is gated behind
--install-depsorINSTALL_DEPS=1. - Node repos use
npm ci --ignore-scriptswhere possible. - Python repos prefer
uv sync,poetry install, orpip install -r requirements.txtdepending on the manifest. - Go repos build locally with
go build ./.... - Manual steps still required: Chrome unpacked extension load, native host install with extension ID, private npm auth for TS repos, internal secrets and service credentials.
Common prerequisites
- git, Node.js 20+, npm
- Bun (for
hmm-aha-search) - Python 3.10+,
uvand/or Poetry - Go (for
alexandriaanddevprod-mcp-router) - Chrome / Chromium
- Internal auth tools (Kanopy OIDC / CorpSecure, private npm auth)
Manual credentials likely needed
- Salesforce integration user credentials / JWT materials
- Atlas API keys or Atlas project-level credentials
- Jira / Confluence tokens
- Glean, Slack, Monday.com, Google Calendar, Gmail, Discord, S3, or Trino credentials depending on scenario
- Private npm auth for
@ts-toolspackages
Glossary
| Term | Definition |
|---|---|
| FTDC | Full Time Diagnostic Data Capture; MongoDB diagnostic metrics bundle used for storage/performance troubleshooting |
| MCP | Model Context Protocol; tool-integration protocol used by devprod-mcp-router and ts-search-explain-helper for agent-accessible tooling |
| MV3 | Chrome Extension Manifest V3, the extension model used by mdb-tam |
| Native Messaging | Chrome mechanism for talking to local executables through a registered host manifest |
| Support API | Backend service surface for automated support case operations |
| Customer Hub | TS customer-context application that centralizes support/operator views |
| Explain plan | Query or Atlas Search execution plan used to diagnose performance issues |
| Aha | Product/feature tracking system; relevant for roadmap/feature-ticket context |
| Kanopy / CorpSecure | Internal platform/auth surfaces commonly needed for service access and binaries |