screen-architecture-spec · supabase_json
Bezel Workflow screen-architecture-spec
screen-architecture-spec artifact · for Bezel Workflow · status draft
screens
| id | name | type | notes | purpose | key data |
|---|---|---|---|---|---|
| workflows | Workflows | core | Auth/login/create account are API-owned and out of this screen spec. | Home base for workflow operations. | - workflow id/name/status - active toggle - last_run_at - last_run_status - failure_count - owner/project |
| template-gallery | Template Gallery | core | Fast start path for real automations. | ||
| workflow-builder | Workflow Builder | core | Human-readable Flow authoring surface. | ||
| step-config-panel | Step / Connector Config Panel | core | This is where Bezel can make advanced agent behavior feel understandable instead of exposing raw agent config. | Configure selected trigger/step without leaving builder. | |
| credentials-connections | Credentials / Connections | core | Manage connector auth and workflow-scoped credential references. | ||
| runs-executions | Runs / Executions | core | All workflow runs across workflows. | ||
| run-detail-timeline | Run Detail / Timeline | core-differentiator | Explain exactly what happened in a workflow run. | ||
| debug-replay | Debug / Replay | core | Recover and iterate from failed production runs. | ||
| agent-pool-roles | Agent Pool / Roles | bezel-differentiator | Show capacity and health of role-based agent pools. | ||
| queue-topic-monitor | Queue / Topic Monitor | bezel-differentiator | Operational view of pull queues/topics. | ||
| self-healing-review | Self-Healing Review | bezel-differentiator | Human review surface for connector drift and intelligent remediation. | ||
| policies-guardrails | Policies / Guardrails | core-differentiator | Policy can begin as workflow-level settings, then expand to workspace/project defaults. | Govern workflow behavior before production activation. |
Agent Handoff
Start Here
screen-architecture-spec artifact · for Bezel Workflow · status draft
Completion Evidence
No explicit evidence field yet. Require tests, screenshots, linked PRs, or reviewed outputs before marking complete.
Artifact Shape
- kind: screen-architecture-spec
- defer: 6 items
- project: Bezel Workflow
- screens: 12 items
- principle: Do not copy n8n screen-for-screen. Use n8n to establish expected workflow-product baseline, then add Bezel’s pull-queue, agent-pool, Router decision, and self-healing surfaces.
- build sequence: 11 items
- research basis: 4 items
- schema version: screen-architecture.v1
- source memory id: 99edd8d2-5780-43a2-bbd0-03982353768a
- next design artifacts: 3 items
- architecture decision ref: 3a1bb081-67f9-41ae-9600-a91fbb14ecda
Structured Payload
Machine-readable source fields
kind
screen-architecture-spec
defer
- Advanced sharing/RBAC beyond API auth/account baseline
- Workflow version history
- Marketplace/community templates
- Billing/usage
- Full audit export
- Custom connector builder UI
project
Bezel Workflow
screens
| id | name | type | notes | purpose | key data |
|---|---|---|---|---|---|
| workflows | Workflows | core | Auth/login/create account are API-owned and out of this screen spec. | Home base for workflow operations. | - workflow id/name/status - active toggle - last_run_at - last_run_status - failure_count - owner/project |
| template-gallery | Template Gallery | core | Fast start path for real automations. | ||
| workflow-builder | Workflow Builder | core | Human-readable Flow authoring surface. | ||
| step-config-panel | Step / Connector Config Panel | core | This is where Bezel can make advanced agent behavior feel understandable instead of exposing raw agent config. | Configure selected trigger/step without leaving builder. | |
| credentials-connections | Credentials / Connections | core | Manage connector auth and workflow-scoped credential references. | ||
| runs-executions | Runs / Executions | core | All workflow runs across workflows. | ||
| run-detail-timeline | Run Detail / Timeline | core-differentiator | Explain exactly what happened in a workflow run. | ||
| debug-replay | Debug / Replay | core | Recover and iterate from failed production runs. | ||
| agent-pool-roles | Agent Pool / Roles | bezel-differentiator | Show capacity and health of role-based agent pools. | ||
| queue-topic-monitor | Queue / Topic Monitor | bezel-differentiator | Operational view of pull queues/topics. | ||
| self-healing-review | Self-Healing Review | bezel-differentiator | Human review surface for connector drift and intelligent remediation. | ||
| policies-guardrails | Policies / Guardrails | core-differentiator | Policy can begin as workflow-level settings, then expand to workspace/project defaults. | Govern workflow behavior before production activation. |
principle
Do not copy n8n screen-for-screen. Use n8n to establish expected workflow-product baseline, then add Bezel’s pull-queue, agent-pool, Router decision, and self-healing surfaces.
build sequence
- Workflows list
- Workflow builder shell
- Step / connector config panel
- Credentials / connections
- Runs / executions list
- Run detail / timeline
- Agent pool / roles
- Queue / topic monitor
- Debug / replay
- Self-healing review
- Policies / guardrails
research basis
- n8n workflows are node-based automations with templates, sharing, executions, credentials, settings, and workflow history.
- n8n distinguishes manual/test executions from production executions.
- n8n supports workflow-level and all-execution lists, retry failed workflows, debug failed executions in editor, pinned data, dirty nodes, and execution data redaction.
- n8n queue mode queues whole workflow execution IDs to workers; Bezel’s product must expose step/topic/agent claim behavior as its differentiator.
schema version
screen-architecture.v1
source memory id
99edd8d2-5780-43a2-bbd0-03982353768a
next design artifacts
- workflow-builder-ux-spec
- run-timeline-ux-spec
- connector-config-ux-spec
architecture decision ref
3a1bb081-67f9-41ae-9600-a91fbb14ecda