Pipeline Hygiene Agent

Adam's take

Forecast meetings are only as honest as the pipeline under them. This agent recommends cleanup rather than doing it, because deleting a rep's deal by robot is how RevOps loses the room.

Required tools

Optional tools

Expected value

Cleaner pipeline; forecast trust; fewer surprise slips

Deployable workflow pack

The full pack includes the agent spec, deployment guide, architecture diagrams, n8n workflow skeleton, Clay template, HubSpot properties, JSON schemas, OpenAPI spec, cost and ROI calculators, demo data, QA checklist, and governance notes.

How to deploy this agent

Every sales leader has sat through a forecast call where a deal "closing this week" turns out to have no next step logged and hasn’t been touched in three weeks. The ten checks that catch this are simple individually; running all ten, on every open deal, every day, by hand, is the part nobody has time for. This agent runs them daily and recommends cleanup — it never closes or re-stages a deal itself.

Core signals

Input schema

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://deploytheagent.com/schemas/agent-009-input.schema.json",
  "title": "Agent #009 Pipeline Hygiene Agent Input",
  "type": "object",
  "required": [
    "record_id"
  ],
  "properties": {
    "record_id": {
      "type": "string",
      "description": "CRM record ID (company/contact/deal depending on agent)."
    },
    "domain": {
      "type": "string"
    },
    "config_version": {
      "type": "string",
      "description": "Versioned config used for this run."
    }
  },
  "additionalProperties": true
}

Output schema

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://deploytheagent.com/schemas/agent-009-output.schema.json",
  "title": "Agent #009 Pipeline Hygiene Agent Output",
  "type": "object",
  "required": [
    "record_id",
    "pipeline_hygiene_score",
    "pipeline_hygiene_flags",
    "pipeline_hygiene_recommendations",
    "pipeline_hygiene_scan_date"
  ],
  "properties": {
    "record_id": {
      "type": "string"
    },
    "pipeline_hygiene_score": {
      "type": "string",
      "description": "Agent #009 output field."
    },
    "pipeline_hygiene_flags": {
      "type": "string",
      "description": "Agent #009 output field."
    },
    "pipeline_hygiene_recommendations": {
      "type": "string",
      "description": "Agent #009 output field."
    },
    "pipeline_hygiene_scan_date": {
      "type": "string",
      "description": "Agent #009 output field."
    }
  },
  "additionalProperties": false
}

Scoring / classification logic

Hygiene score starts at 100 and subtracts points per failed check, floored at 0. Every flagged issue comes with a specific, generated recommendation — "Close date is 4 days in the past, confirm new close date or update forecast category" — not a generic "clean up this deal" nudge.

HubSpot CRM properties

n8n workflow skeleton

The Pipeline Hygiene Agent scans every open opportunity for the ten checks that quietly break forecast trust — missing amount, missing or past close date, no next step, stale activity, no champion — and recommends cleanup actions. It never closes or re-stages a deal itself. Configure every node marked CONFIGURE before enabling the trigger.

1. Trigger
   CONFIGURE: set cadence per agent-spec.md's What It Does section. Leave disabled until Step 6 of deployment-guide.md.
2. Load versioned config
   CONFIGURE: replace with your versioned config built from real data (see deployment-guide.md Step 1).
3. Fetch source data
   CONFIGURE: {{HUBSPOT_CREDENTIAL}}. Pull the fields listed in agent-spec.md's Inputs table.
4. Validate + gate records
   Drops records missing required inputs or failing suppression/gate checks defined in agent-spec.md. Logs every drop.
5. Apply workflow logic
   Implements the Workflow Logic table from agent-spec.md. Writes model/config version alongside every derived value for auditability.
6. Write agent-owned CRM fields
   CONFIGURE: {{HUBSPOT_CREDENTIAL}}. Writes only: pipeline_hygiene_score, pipeline_hygiene_flags, pipeline_hygiene_recommendations, pipeline_hygiene_scan_date. Never writes sales-owned fields.
7. Route to human review / alert
   CONFIGURE: {{SLACK_CHANNEL}}. Fires for records requiring human review per agent-spec.md's Governance section.

Cost calculator

ROI calculator

Governance notes

• This agent only recommends — it has no CRM write permission for stage, amount, or close-date fields, only for its own agent-owned flag and score fields.
• Closing or re-staging a deal is always a human action; the agent’s job ends at a specific, actionable flag.
• Stage-age thresholds are set per stage in config, since normal time in Discovery and normal time in Negotiation are very different numbers.
• The champion check degrades gracefully when the Buying Committee Mapper isn’t deployed — it’s disabled in config rather than always failing.

Failure modes

• Agent accidentally closes or re-stages a deal: prevented structurally — write credentials are scoped only to agent-owned fields, and a QA test asserts no write attempt ever reaches stage, amount, or close-date fields.
• Vague next steps not caught: the vague-phrase pattern list is versioned and expanded every time a miss gets caught in review.
• Stage-age threshold flags normal early-stage deals: prevented by requiring a threshold per stage in config rather than one global number.
• Forecast-inconsistency check never fires: a deliberately inconsistent test deal (Commit category, 90-day-out close date) is part of QA specifically to verify this check actually catches what it’s supposed to.

QA checklist

FAQ

Related agents