RFC-WF-0023
Template UX Patterns for Conversational Ops (TUXP)
Status: Draft Standard
Version: 1.0.0
Date: 20 Nov 2025
Category: Standards Track
Author: FullAgenticStack Initiative
Dependencies: RFC-WF-0002 (WWCS), RFC-WF-0003 (CCP), RFC-WF-0005 (CRCD), RFC-WF-0007 (OoC), RFC-WF-0008 (RCP), RFC-WF-0015 (PPGP), RFC-WF-0018 (SMCL)
License: Open Specification (Public, Royalty-Free)
Abstract
This document specifies Template UX Patterns for Conversational Ops (TUXP) for WhatsApp-first systems. TUXP defines a set of normative response and interaction templates to ensure consistent, safe, and efficient conversational operations across command execution, disambiguation, confirmation, observability drill-down, and recovery workflows. TUXP standardizes message structures (summary-first), numbered actions, preview blocks, error/denial explanations, and progressive disclosure patterns, enabling consistent UX across implementations while preserving protocol correctness.
Index Terms— conversational UX, templates, drill-down, disambiguation, confirmations, recovery flows, WhatsApp-first.
I. Introduction
Even with correct protocols (CCP/ACSM/IDS/EAS), WhatsApp-first systems can fail UX-wise: ambiguous prompts, missing previews, inconsistent menus, and verbose dumps that break operator workflow. TUXP provides standardized templates that:
- reduce ambiguity and accidental execution
- improve learnability (consistent patterns)
- enable tooling to generate predictable conversational UI
- align with WWCS conversion patterns (web → WhatsApp)
TUXP is not about branding tone; it’s about operational clarity and safety.
II. Scope
TUXP specifies:
- Required template set and when each MUST be used
- Formatting rules (summary-first, numbered actions, stable identifiers)
- Disambiguation patterns for ambiguous intents/targets
- Confirmation/approval preview structures
- OoC drill-down patterns for evidence and errors
- Recovery flow patterns (RCP options and execution)
- Policy-driven verbosity and redaction constraints (PPGP)
TUXP does not define WhatsApp UI widgets; it defines textual structures compatible with WhatsApp constraints.
III. Normative Language
MUST, MUST NOT, SHOULD, SHOULD NOT, MAY are normative.
IV. Definitions
Template: A structured response pattern with required fields.
Summary-first: A response that begins with the minimum actionable summary before details.
Drill-down: Progressive disclosure via numbered choices or explicit follow-up commands.
Stable Identifier: A reference token like command_id, correlation_id, or resource id.
V. Design Goals
TUXP MUST ensure:
- G1. Safety: reduce accidental destructive actions.
- G2. Consistency: same situation → same response structure.
- G3. Efficiency: minimal text overhead; drill-down for details.
- G4. Traceability: include stable identifiers for follow-ups.
- G5. Policy Compliance: obey redaction/detail policies.
VI. Core Formatting Rules (Normative)
A. Summary-First Rule
Any response about a command or system state MUST start with:
- what happened / what will happen
- the key identifier(s)
- next available actions (if any)
B. Numbered Action Rule
Whenever multiple actions are possible, the system MUST present options as:
1 - <action>2 - <action>- …
Options MUST be stable for the current context.
C. Preview Block Rule
Before executing any mutation requiring confirmation/approval, the system MUST show a preview block containing:
- entity + action + target id
- impact summary
- reversibility / recovery hint (if applicable)
- required confirmation method (token/phrase)
D. Identifier Rule
Responses SHOULD include command_id (or a reference) when applicable, enabling OoC/RCP follow-ups.
E. Redaction Rule
Sensitive fields MUST be masked/dropped according to policy; templates MUST not “accidentally print” secrets or full PII.
VII. Required Template Set
T1 — Command Acknowledgement (Post-Canonicalization)
When: after CCP envelope creation (canonicalized).
MUST include: command_id, summary, next step (confirm/authz).
Structure (normative fields):
- Summary line
-
command_idline - Next step line (confirm required? waiting authorization?)
T2 — Disambiguation List
When: ambiguous target/intent.
MUST include: numbered candidates with stable ids.
Structure:
- “Ambiguity detected” summary
- Candidate list with ids
- Instruction: “reply with number” (or explicit id)
T3 — Confirmation Request (Mutation)
When: mutating S1 commands.
MUST include: preview block + confirmation instruction.
Structure:
- Preview block
- Confirmation method (explicit phrase or yes/no per policy)
- Cancel option
T4 — Strengthened Confirmation (Destructive)
When: S2 destructive commands.
MUST include: preview + token/phrase binding + expiry.
Structure:
- Preview block
- “Confirm with TOKEN/PHRASE bound to target”
- Expiry time
- “Deny/Cancel” option
T5 — Approval Request (HITL)
When: approvals required (HAPF).
MUST include: requester, risk class, preview, approve/deny choices, expiry.
T6 — Execution Result
When: command reaches terminal (executed|rejected|failed|canceled|compensated).
MUST include: status, summary, affected resources (if non-sensitive), next actions.
T7 — Denial / Rejection Explanation
When: rejected by policy/authz/ambiguity.
MUST include: reason category and minimal remediation (“missing scope”, “needs approval”, “choose target”).
T8 — OoC Status Summary
When: OoC status queries.
MUST include: summary, lifecycle state, last error (if any), drill-down options.
T9 — Evidence Drill-down
When: user requests evidence.
MUST include: list of artifacts (types + timestamps), with numbered drill-down to view a specific artifact (redacted).
T10 — Recovery Options (RCP.OPTIONS)
When: a command is failed/blocked/partially executed.
MUST include: current state, why recoverable, numbered list of recovery actions with risk labels.
T11 — Recovery Execution Preview
When: user selects retry/reprocess/compensate.
MUST include: target command id, risk, expected effect, confirmation/approval method required.
VIII. Policy-Driven Variations (PPGP Integration)
Templates MUST be parameterized by governance policies, including:
- max detail level (OoC)
- redaction strategy
- confirmation/approval method
- verbosity caps under RAOS degradation modes
Implementations MUST be able to switch profiles without breaking template structure.
IX. Lifecycle Alignment (SMCL)
Templates SHOULD align with lifecycle stages:
-
canonicalized→ T1 -
confirmation_required→ T3/T4 -
approval_required→ T5 -
started→ short progress update (optional) - terminal states → T6/T7
- OoC queries → T8/T9
- recovery flows → T10/T11
X. Conformance Requirements
An implementation is TUXP-compliant if:
- it uses numbered actions for multi-choice interactions
- it shows preview blocks before mutating executions requiring confirmation/approval
- it uses disambiguation lists when needed
- it presents OoC and RCP outputs summary-first with drill-down
- it respects redaction/detail policies in all templates
XI. Security Considerations
Bad UX is a security vulnerability in conversational systems: ambiguous confirmations and verbose leakage cause real harm. Templates MUST be designed to prevent “accidental admin” and “leaky observability” scenarios. Under abuse conditions, systems SHOULD degrade to safer templates (summary-only, stricter confirmation).
XII. Conclusion
TUXP standardizes the conversational “operational UI kit” for WhatsApp-first: predictable templates for execution, safety, observability, and recovery. This reduces ambiguity, increases safety, and enables tooling to generate consistent conversational experiences across systems—without sacrificing protocol rigor.
References
[1] RFC-WF-0002, Web-to-WhatsApp Conversion Standard (WWCS).
[2] RFC-WF-0003, Conversational Command Protocol (CCP).
[3] RFC-WF-0005, Command Registry & Capability Declaration (CRCD).
[4] RFC-WF-0007, Observability over Conversation (OoC).
[5] RFC-WF-0008, Recovery & Compensation Protocol (RCP).
[6] RFC-WF-0015, Policy Packs & Governance Profiles (PPGP).
[7] RFC-WF-0018, State Model & Command Lifecycle (SMCL).
Concepts and Technologies
Summary-first UX, numbered action menus, disambiguation lists, preview blocks, strengthened confirmations, approval prompts, evidence drill-down, recovery option menus, policy-driven verbosity/redaction, lifecycle-aligned conversational UI patterns.
Top comments (0)