Telynt / Documentation

Search documentation

Search platform contracts, developer guides, package manuals and the Technology Consulting & Delivery solution.

Browse documentation

Decisions, Commitments, Actions, and Follow-Up

Decisions, Commitments, Actions, and Follow-Up — Technical reference

On this page

This reference owns the implementation contracts of BND-WORK. Read the package guide for usage, configuration, dependencies and upgrades. Generic execution, isolation and admission remain governed by Platform Documentation.

Data model and interfaces

DataModelPackage: DM-WORK

Owned entities: decision, proposal, commitment, acknowledgement, acceptance, rejection, action, open question, follow-up, external-write draft.

InterfacePackage contributions: IF-WORK-MANAGEMENT.

Each business object and source reference remains tenant- and purpose-scoped. The package owns schema evolution and declared migration behavior; the platform enforces schema registration, authorization and versioned installation. Consumers reference this model rather than creating competing canonical copies.

Workflows

WorkflowPackage: WFP-WORK-LIFECYCLE

  • Decision and Commitment Extraction
  • Action and Open Question Extraction
  • Owner and Date Resolution
  • Action Lifecycle
  • Commitment Fulfilment
  • Follow-Up Drafting
  • Reminder Scheduling
  • External Write Approval
  • Compensation and Remediation

These are named definitions within the workflow package, not separate invented workflow IDs. Workflow runs resolve package versions, actor and purpose, authorized inputs, retries, review gates and output schemas through the environment lock. Temporal provides durable execution and Kafka carries declared lifecycle/domain events.

Policies, evaluations and tools

POL-CONSEQUENTIAL-ACTION, EVAL-DECISION-ACTION, task, calendar, messaging, CRM, project, and document draft tools.

Qualification covers the selected model/runtime profile, languages, access scope, failure handling and the package's domain-specific outcomes. Tool execution uses the platform gateway; package code cannot increase its own permissions or write directly to external systems.

Owned agents

M02 — Decision, Commitment, and Acceptance Agent

Agent ID: effority.decision-commitment

Purpose: Extracts decisions, proposals, commitments, acknowledgements, operational acceptances, rejections, and possible contractual acceptances.

Triggers: Topic structure completion, transcript correction, document update, or related review outcome.

Inputs: Topic windows, speakers, dates, business objects, earlier commitments, applicable documents, and authority context.

Outputs: Structured decision and commitment artifacts with actor, recipient, object, due or effective date, evidence, status, and formalization need.

Tools: evidence.search, graph.read, finding.read, document.clause_read, and artifact.derive.

Review and limits: Possible contractual acceptance always requires competent human review. The agent never treats silence or vague agreement as binding acceptance.

Primary metrics: Precision and recall by artifact type, owner/date extraction accuracy, acceptance false-positive rate, and citation coverage.

M03 — Action and Open Question Agent

Agent ID: effority.action-question

Purpose: Extracts executable actions and unresolved questions from meetings and related communications.

Triggers: Topic structure completion, follow-up message, review modification, or source correction.

Inputs: Topic windows, participant roles, previous actions, due dates, dependencies, and current workflow state.

Outputs: Action, owner, due date, dependency, completion condition, question, expected respondent, and overdue or answered state.

Tools: finding.read, workflow.read, directory.read, calendar.read, and artifact.derive.

Review and limits: Missing owners or dates remain explicit gaps. The agent never assigns responsibility solely from hierarchy when it was not stated or configured.

Primary metrics: Action coverage, owner accuracy, date accuracy, open-question precision, and overdue-state accuracy.

M07 — Follow-Up Agent

Agent ID: effority.follow-up

Purpose: Converts validated meeting outcomes into reviewable follow-up communications, reminders, and internal workflow drafts.

Triggers: Meeting analysis completion, review completion, due-date threshold, or user request.

Inputs: Validated minutes, decisions, actions, commitments, recipients, communication template, due dates, and approval matrix.

Outputs: Follow-up email draft, reminder, meeting invitation draft, task draft, risk draft, decision record draft, or escalation draft.

Tools: artifact.read_validated, template.read, calendar.draft, message.draft, workflow.draft, and approval.request.

Review and limits: Every external communication and system change remains a draft until approval.

Primary metrics: Draft acceptance rate, edit distance, reminder timeliness, missing-recipient rate, and unauthorized-write attempts.

U06 — External Synchronization Draft Agent

Agent ID: effority.external-sync-draft

Purpose: Converts reviewed Telynt artifacts into schema-valid drafts for CRM, project, ticketing, calendar, messaging, document, and BI systems.

Triggers: Approved finding, explicit user request, workflow step, or scheduled synchronization review.

Inputs: Reviewed artifacts, destination schema, current destination record, mapping rules, conflict policy, approval matrix, and idempotency state.

Outputs: Proposed create or update payload, field-level diff, evidence references, validation result, conflict report, rollback or compensation information, and approval request.

Tools: connector.schema_read, destination.record_read, mapping.read, external-sync.draft, and approval.request.

Review and limits: It never performs the external write. The workflow engine executes an APPROVED_WRITE only after a recorded authorized approval.

Primary metrics: Schema-valid draft rate, field-mapping accuracy, approval rate, edit distance, conflict detection, duplicate prevention, and rollback success.

Implementation profiles

The package composes the data, workflows, policies and interfaces above using the shared runtime. The owned-agent contracts specify inputs, outputs, tool boundaries, review requirements and measurable outcomes. Deterministic platform operations are not represented as additional agents.

Integration boundaries

A rollback of software does not undo an email or an external write. Such effects need explicit compensation or remediation.

Owned screen contracts: SCR-MTG-04, SCR-WRK-01, SCR-WRK-02.

See solution workflows for cross-package behavior. Screens, agents and workflows keep one canonical owner even when other bundles invoke or enrich them.