Value-First AI seat
Operations
Loom
HubSpot Automation / Workflow Write Gateway
Loom is an AI seat of the Value-First Team — software with a defined role on the team, not a person.
Schedules and manages autonomous background tasks
About Loom
Loom is the organization's HubSpot automation gateway — the single disciplined writer of workflows, peer to Ledger (CRM) and Broadcast (Marketing and Content). She weaves the org's durable automation directly in HubSpot: every workflow validated against the live flows API, its blast radius named before anything enrolls a live record, read back and verified after. Her craft was forged by a hard lesson — recurring work that fires into a void, where nobody receives the result, is motion, not value. A HubSpot workflow is the durable answer a homegrown scheduler never was: it lives in the platform, it is observable, and it enrolls a real receiver. Loom makes sure every automation she ships is safe to re-run and actually lands.
Loom
Background Worker Orchestrator
Schedules and manages autonomous background tasks
“"Background intelligence is the foundation of informed decisions. But execution must be on-demand, not fire-and-forget."”
Identity
Loom orchestrates background worker scripts -- health checks, intelligence scripts, and recurring operations. Worker scripts live in agents/background-workers/workers/ and run on-demand via npx tsx when slash commands need fresh data. The previous GitHub Actions execution path was replaced with inline execution during operational commands like /daily-ops and /all-hands.
Current State
An honest assessment of where this agent stands today.
What Works
- Worker registry tracking all available scripts
- On-demand execution via npx tsx for slash commands
- BU health check scripts for all 6 business units
What Doesn’t Work Yet
- No automated scheduling -- all execution is inline during slash commands
- Previous GitHub Actions path deactivated
Leadership Commentary
“Loom represents the shift from always-on background daemons to on-demand inline execution. The workers exist and produce valuable data, but they run when I need them during /daily-ops or /all-hands rather than on a fixed schedule. This is intentional -- on-demand execution is more reliable than hoping a cron job ran.”
Delegation Contract
The observable, falsifiable standard this agent is held to.
Quality Bar
Worker registry accurately reflects available scripts, and on-demand execution produces valid output.
- Worker registry lists all available scripts with metadata
- Each worker script runs successfully via npx tsx
- BU health check scripts produce valid JSON output
- No forbidden language
Invocation Triggers
Feedback Loop
Worker failures: when scripts fail, error output feeds to Squire for dependency investigation and Mender for automated resolution.
Handoff
All slash commands that need fresh data (V runs workers inline)
Scope Boundary
Loom orchestrates workers. Individual workers have their own domain logic (Pulse for health, Sentinel for relationships, etc.).
The rest of the team
Other Value-First AI seats in Operations.
The team behind the work is on the record.
Every seat on the Value-First Team has a defined role, a public standard, and a name you can look up.