CiterLabsFREE PRACTICAL RESOURCE / NO SIGNUP
Better lead intake.
Follow-up you can review.
An actionable automation kit for small service businesses. Map the workflow, adapt the templates and test safely before connecting real customer data.
Version 0.1.0 · Synthetic examples, not client results
WHAT THIS HELPS YOU DO
From scattered enquiries
to a consistent next step.
Start with one inbox or form and one record system. Prepare structured records and response drafts, while keeping approval, exceptions and follow-up decisions visible.
This deterministic rules-based planner does not use a model. This kit does not configure a CRM or send messages. It gives you a concrete plan and reusable materials to adapt to your own tools.
FOR PEOPLE
Plan it in your browser.
Enter assumptions below and download an editable blueprint. Everything calculates on your device.
Use the planner ↓FOR AGENTS / LIVE A2A SERVICE
The same function, through A2A.
A real A2A 1.0 endpoint uses the official SDK to return the same deterministic plan. Public discovery and a synthetic request have been verified with the official SDK.
Use the agent interface ↗THE WORKFLOW
Seven steps.
Clear review gates.
Begin in draft-only mode. Treat every enquiry as untrusted input; customer-facing actions need explicit permission.
- 01
Capture
Record source, request ID, received time, requested service, plain-language need and owner. Missing stable IDs go to manual exception review. Keep an approved minimal field set.
Review gate: Do not ingest private attachments or sensitive records into unapproved tools.
- 02
Normalize and deduplicate
Normalize fields; use source + request ID as the stable deduplication key. Missing fields route to manual review.
Review gate: A repeat event must not create another lead or draft.
- 03
Triage
Match a request against the approved service scope. Flag urgent, ambiguous, sensitive or out-of-scope enquiries.
Review gate: Never infer identity, protected traits, budget or permission to contact.
- 04
Prepare a draft
Use approved business facts and the response template. Ask for missing scope details. Link every factual business claim to the approved source.
Review gate: No invented price, availability, promise or testimonial. Draft preparation only.
- 05
Human approval
An owner checks recipient, facts, tone, next step and any contact preferences. Approve, edit or reject.
Review gate: Sending requires an explicit approved action; use a separate permission boundary.
- 06
Record and follow up
After an approved send, record status and a reviewer-selected follow-up date. Prepare a follow-up draft only where contact is appropriate.
Review gate: Stop on reply, decline or opt-out, including later replay events. Never treat this kit as consent for marketing.
- 07
Handle exceptions
Use a visible exception queue; cap retries; log the failure without unnecessary message content. Owner reviews before restart.
Review gate: Pause sending on connector errors or uncertain send status; never blindly resend.
REUSABLE MATERIALS
Download. Edit.
Make it fit your work.
Plain text and CSV files work in your existing editor or spreadsheet. No account, tracking or private data required. When adapting the tracker, escape spreadsheet cells that begin with =, +, - or @; untrusted lead text must not become a formula.
Editable workflow and operating checklist
TXT / editable textIntake, acknowledgement and follow-up templates
TXT / editable textEmpty editable lead tracker
CSV / spreadsheetSix synthetic acceptance scenarios
CSV / spreadsheetPilot acceptance checklist
CSV / spreadsheetReproducible community contribution template
TXT / editable textSix templates. Clear reuse permission.
The original contents of the six files listed above are licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). You may copy and adapt them, including commercially, with attribution. Credit CiterLabs — Lead intake and follow-up kit, link to this source page and the license, and indicate changes. Do not imply endorsement.
This grant covers only workflow-blueprint.txt, response-templates.txt, lead-tracker.csv, synthetic-enquiries.csv, acceptance-checklist.csv and community-contribution-template.txt. Generated blueprint JSON, this resource JSON, schemas, implementation and website code, branding, imagery and third-party material are excluded. Compliant reuse permissions are irrevocable. Check any separate data or third-party rights before implementation.
1. Define the smallest useful scope
Choose one enquiry source, one service type and a reviewer. Fill in the owner, approved facts, permitted fields and stop conditions in the blueprint. Avoid collecting more information than the next decision needs.
2. Choose tools you already control
A shared inbox and spreadsheet can be enough for the first test. If using an automation platform or model, check permissions, retention, costs and data handling before connecting it. This kit does not require a paid tool or a model.
3. Build draft-only stages
Capture and deduplicate first. Then prepare a structured record and draft from approved facts. Keep the send action disabled. The optional drafting instruction is an example, not protection by itself: enforce tool permissions outside the model and review every output.
4. Run the acceptance scenarios
Use the six synthetic cases: normal, duplicate, missing scope, unsupported promise, embedded instruction and stop request. Also simulate connector failure, uncertain send status and the owner pause switch. Record observed behavior in the acceptance checklist; a failed check stays a blocker.
5. Measure before expanding
Observe manual handling time and draft plus review time on comparable tasks. Include corrections and maintenance. A limited real-data pilot requires approved access, a reviewer, explicit sending permissions and an exception plan. Recheck assumptions before extending the scope.
YOUR ASSUMPTIONS / YOUR SCENARIO
Is the effort
worth investigating?
Enter your own figures. The calculator compares monthly handling effort and estimates when setup effort would be recovered as capacity. It does not predict revenue or guarantee savings.
How it works
Unscoped enquiries retain your current handling time. Scoped enquiries use your proposed draft + review time. Weekly enquiries are converted to a monthly average using 52 ÷ 12. Monthly maintenance is added to the proposed effort. The hourly-cost assumption is included in the downloadable JSON as a theoretical capacity valuation, not realized cash savings.
Capacity difference = current effort − proposed effort. Setup recovery = setup hours ÷ positive monthly capacity difference. If there is no positive difference, no recovery is shown.
Excludes tool/model spend, demand changes, approval delays and quality effects. Verify real effort and quality before making a business decision.
BUILD USEFUL RESOURCES TOGETHER
What would make
this kit more useful?
Share a suggestion, a non-confidential workflow outline or a synthetic test case. Tell us what you tried, which step was unclear and what result you observed. Keep customer records and confidential material out of the email.
Share feedback by email ↗Email participation is available. There is no live forum or membership program, and no response-time promise.
FOR DEVELOPERS AND AGENTS
Use the same public resource data.
Read the workflow, asset URLs, estimator formulas and synthetic examples in JSON. Respect the stated reuse status and data boundary.
Resource JSON ↗JSON Schema ↗Use the live A2A service ↗The JSON links are static resource data. A separate executable A2A 1.0 service is live at /a2a. Discover it at /.well-known/agent-card.json; use the exact paths without query strings.