agentic-development-platform

Decompose a feature or intent into small, context-bound backlog items in the SQL backlog, for a planning model (Opus or Sonnet). Use before a swarm of workers: turn one feature into tightly scoped, logged items, each with a write scope and a context profile, so cheaper sub-agents can each work one item in isolation and total token use drops.

Orchestrate a context-bound, backlog-logged multi-agent swarm to cut token use. Use to run a feature as a planner (Opus or Sonnet) that logs small backlog items, then cheaper workers (Sonnet or Haiku) that each implement one item under its bounded context packet, while the parent keeps integration, the SQL backlog, the audit log, and verification.

Select, record, and validate ADG evidence. Use when choosing targeted checks, recording failed commands, linking tests to requirements, writing audit events, registering advisory gaps, or preparing feature/release evidence.

Elicit generic ADG feature intent into a normalized requirements graph. Use when naming a feature, defining how it fits the platform, deriving RBAC-aligned stories and use cases, or creating success, anti-success, scenario, and gap evidence in config/agentic/elicitation.json.

Work a single backlog item under a bounded context packet, for a context-cheap sub-agent (Sonnet or Haiku). Use when an orchestrator hands you one item id to implement in isolation: claim it, load only its context:item packet, edit only its write scope, record evidence, and complete it. Keeps token use low and work traceable.

Use for small UI, copy, docs, styling, or visual polish fixes that should avoid full governance unless risk increases.

Use when work needs release, RC, GA, verified, release-ready, or signed-off claims backed by traceable evidence.

Maintain the append-only audit log and SQL-first backlog of a repo that uses the Agentic Development Platform. Use when working on feature delivery, bug fixes, route or scope changes, integration status, release evidence, comments, decisions, or status updates that should be recorded in data/audit/audit-log.jsonl and the data/backlog.sqlite tracker.

Use when a PDF, runbook, or signoff document must be checked against the repo's actual commands, schema, scripts, and gates.

Use before implementation or review work to choose the cheapest safe Proofline lane and avoid unnecessary governance gates.

Use for read-only exploration spikes where the user wants options, diagnosis, or a recommendation without full governance.

Validate requirements-to-UX-as-code lineage. Use when features, user stories, use cases, requirements, scenarios, journeys, surface states, RBAC roles, or UX contracts are added or changed.

Guide an agent through bounded feature-slice delivery. Use before implementation work that needs reduced rework, scoped context, stop conditions, evidence, and targeted checks.

Maintain standards-linked governance evidence. Use when controls, evidence, maturity scores, security practices, AI governance, privacy, or deliverable auditability changes.

Create or maintain compact bord.room agent skills. Use when adding, revising, validating, or promoting SKILL.md packages for context selection, UX contracts, surface maps, RBAC, data, testing, IaC, docs, release evidence, or agent workflows.

Token-efficient context selection for bord.room V4.1. Use before broad product, UX, architecture, route, RBAC, data, IaC, test, release, or process work to choose bounded SQL-backed files, risks, checks, and stop conditions.

Govern bord.room V4.1 data handling. Use when adding or changing data classes, privacy requests, retention, deletion, export, legal hold, subprocessors, document governance, integration data, storage, or sensitive operational metadata.

Guard bord.room V4.1 database schema, migrations, RLS, repository predicates, tenant context, seeds, and runtime DB role behavior. Use when adding tables, changing repositories, migrations, RLS policies, or tenant/business-owned data access.

Maintain bord.room V4.1 design-system consistency. Use when changing tokens, primitives, page layouts, accessibility, responsive behavior, visual hierarchy, icons, density, tables, cards, forms, or dashboard surfaces.

Reconcile bord.room V4.1 documentation drift. Use when generated docs, source docs, route matrices, tracker mirrors, UX/architecture status, integration lifecycle status, or release planning artifacts may disagree with code or SQL.

Create or update bord.room V4.1 experience contracts. Use when a user goal, workflow, page, primary CTA, server action, route, or launch-critical journey needs explicit persona, authority, state, data, audit, and test coverage.

Review bord.room V4.1 infrastructure and deployment readiness. Use when Terraform, Vercel, Neon, Cloudflare, AWS KMS, storage, environment variables, secrets, regions, workers, rollback, or provider evidence changes.

Govern bord.room V4.1 interaction grammar. Use when adding or reviewing controls, forms, modals, drawers, validation, dirty guards, destructive actions, toasts, keyboard behavior, focus handling, loading, retry, or mobile interaction states.

Audit bord.room V4.1 RBAC, entitlement, tenant scope, and business scope. Use when protected routes, actions, queries, billing/package behavior, platform admin, support access, or persona workflow evidence changes.

Curate bord.room V4.1 launch and release evidence. Use for go/no-go packets, blocker counts, deferrals, owners, risk, rollback, restore proof, cost review, security evidence, support readiness, and release-gate decisions.

Map bord.room V4.1 surfaces. Use when routes, entities, navigation, command search, page titles, hidden/future modules, live states, mobile/desktop states, or entity lifecycle surfaces must be made truthful and queryable.

Maintain bord.room V4.1 technical contracts. Use when APIs, server actions, webhooks, events, integrations, OpenAPI, Zod schemas, payload validation, provider lifecycle status, or typed client behavior changes.

Select and record bord.room V4.1 test evidence. Use when deciding targeted checks, UAT/browser proof, backlog verification, failed-command evidence, traceability gates, or launch evidence for a feature slice.

Low-token bord.room V4.1 feature-slice delivery. Use for normal complete-dev implementation where the SQL backlog remains canonical but full traceability/prepush gates are reserved for checkpoints.

Concise SQL-first traceability for bord.room V4.1 Lite feature-slice delivery. Use to keep evidence truthful without micro-event overhead.

Model cybersecurity controls, abuse paths, prompt-injection defenses, unsafe tool requests, dependency risks, secret handling, and negative security tests as code. Use when security policy, evals, guardrails, auth, secrets, or risky tool behavior changes.

Define and verify data governance as code. Use when data classes, retention, deletion, export, residency, subprocessors, sensitive metadata, privacy behavior, or data-handling evidence changes.

Create or update an ADG experience contract as the primary agent build document. Use when a user goal, workflow, route, action, UI state, RBAC boundary, failure mode, or test-first UX spec needs to be explicit and traceable.

Assess ADG maturity as code. Use when defining 1.0-5.0 scoring, evaluating required domains, validating evidence, producing maturity scorecards, or raising remediation gaps for domains below target.

Model and verify RBAC, entitlement, tenant scope, business scope, and privileged support boundaries as code. Use when protected behavior, role matrices, permission checks, denial paths, or access evidence changes.

Design governed autonomous runtime readiness for ADG. Use when defining capability-scoped execution, sandbox policy, runtime traces, model/tool call records, adapters, or compatibility with future native agent sandboxing.

Map generic ADG product surfaces as code. Use when routes, actions, navigation, hidden or future surfaces, direct URL behavior, page states, or entity lifecycle actions need to be truthful, queryable, and linked to experience contracts.

Execute backlog items from the SQL-first backlog through claim, implementation, tests, traceability updates, and commit-readiness, in any repo using the Agentic Development Platform. Use when asked to build, pick up, run, complete, verify, or orchestrate work from data/backlog.sqlite, especially with multi-agent backlog execution.