- Home
- Skills
- Business & Operations
- Automation Process Hunter — Find Work to Automate
Works with the AI Agents you already use
Automation Process Hunter — Find Work to Automate
Discover real work suited to automation, then prioritize it using process evidence, AI fit, risk, economics, and implementation readiness.
$9.99
Automation Process Hunter — Find Work to Automate
Example session with this skill installed
We process customer refund requests. Staff read emails, check order data, decide whether policy criteria are met, and issue refunds through our store platform. We have no volume or exception-rate figures, and no verified API information. Identify candidate automations, evidence gaps, risks, and a safe pilot. Do not assume integrations or savings.
- Read your context and instructions
- Compiled the automation process hunter
Refund workflow automation discovery
Evidence and unknowns
Known from the request: staff read refund emails, check order data, compare requests with policy criteria, and issue refunds in a store platform.
Unknown: request volume, handling time, eligibility rules, exception and dispute rates, refund limits, fraud exposure, system capabilities/API access, audit logs, access controls, customer impact, and current measured outcomes. No ROI or time savings can be calculated.
Process map to validate
- Receive and classify request.
- Identify order and verify evidence.
- Apply the current refund policy.
- Resolve missing information or exceptions.
- Authorized staff issue or decline refund.
- Notify customer and record outcome. Confirm every step and owner against the actual process.
Candidate work
| Candidate | Potential approach | Gate / human control |
|---|---|---|
| Intake fields and request categorization | Standard form or suggested classification | Review ambiguous, urgent, or sensitive cases |
| Order lookup and policy checklist | Deterministic validation if rules and system access are verified | Confirm data and permissions; never infer eligibility from missing records |
| Draft response | AI-assisted draft using approved policy and verified facts | Human review before customer communication |
| Refund execution | Keep authorized staff approval during discovery and pilot | Verify policy, limits, identity, audit trail, fraud controls, and recovery first |
Safe pilot
Use a permissioned, representative historical sample with outcomes reviewed by policy owners. In shadow mode, compare suggested categorization and checklist results with staff decisions; do not issue refunds or send messages. Measure false eligibility, missed exceptions, reviewer time, disagreement rate, and customer-risk cases. Set thresholds and a stop condition before testing.
Next steps
Confirm process owner and policy version; collect privacy-safe samples and baseline measures; map access and integration evidence; classify exceptions; review controls with finance and support; then decide whether a bounded assisted pilot is justified.
Connects securely to your tools. The creator never sees your data.
About this skill
What this skill does
Automation Process Hunter discovers and evaluates business work worth automating. It examines real processes, finds repetitive or expensive friction, and separates AI-suitable tasks from deterministic automation and human judgment.
Discovery method
It inventories work from authorized SOPs, process maps, interviews, tickets, spreadsheets, workflow logs, and other evidence; maps triggers, tasks, decisions, handoffs, inputs, outputs, exceptions, systems, and owners; and distinguishes root causes from symptoms. Candidates are scored transparently for frequency, volume, stability, input quality, exception burden, time and error costs, customer impact, risk, data readiness, integration readiness, and change effort.
Deliverables
Expect evidence-backed opportunity cards, a prioritized portfolio, current- and future-state workflow concepts, a human-oversight plan, dependencies, and a bounded pilot with measures and acceptance criteria. It recommends elimination, simplification, standardization, deterministic automation, AI augmentation, or keeping work human as appropriate.
Limits
It never invents company processes, employee activity, metrics, system access, integrations, compliance, or ROI. Use only available and authorized sources. Estimates are conditional, and high-consequence decisions stay with accountable people.
Findings are limited by the quality and coverage of authorized evidence. Missing process, volume, cost, outcome, system, or integration data remains unknown. Prioritization is not proof of ROI or deployment feasibility. Validate with process owners and a bounded pilot; do not remove required human judgment or approvals.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago
Permissions
File Scopes
Read only authorized business, process, and workflow evidence supplied or approved by the user. Write requested plans or reports only to the designated workspace. Use external research only when enabled and authorized. Never infer or expose employee-level activity, invent metrics or integrations, or execute workflows. Keep consequential financial, legal, customer, and personnel decisions with authorized humans.