jonathan-vella/apex/apex-workflow-engine
Claude Code and agent skill by jonathan-vella/apex
- community
Checking that the SKILL.md file is still in the repository.
**UTILITY SKILL** — Machine-readable workflow DAG for the multi-step agent pipeline. Defines node types, edge conditions, gates, and fan-out patterns. WHEN: "orchestrator step routing", "resume from graph", "workflow validation", "workflow DAG", "workflow gate", "fan-out pattern". DO NOT USE FOR: Azure infrastructure, code generation, troubleshooting.
Install
npx vskill@latest install jonathan-vella/apex/apex-workflow-engineSKILL.md
Workflow Engine Skill
Provides a declarative, machine-readable workflow graph that the Orchestrator reads instead of relying on hardcoded step logic.
When to Use
- Orchestrator determining the next step after a gate
- Resuming a workflow via
apex-recall show <project> --json - Validating that all steps have proper dependencies and outputs
- Understanding fan-out (parallel sub-steps) and conditional routing
Rules
- Forward DAG —
edgesmust be acyclic; refinement routes are declared separately inreturn_edges. - Source of truth is `templates/workflow-graph.json` — the orchestrator reads this directly; do not encode workflow logic in agent prose
- Gates are blocking — a
gatenode halts downstream execution until human approval is recorded in session state - IaC routing —
step-4is the shared planner; select CodeGen and Deploy nodes usingdecisions.iac_tool. - Fan-out is logical work structure, not permission to invoke main agents or unavailable workers.
The owning main agent handles its outputs; parallelize only independent, authorized work.
- Edge conditions — follow the graph's declared conditions, including refinement returns; the schema allows a string or array.
- Schema evolution — bump
metadata.versionand followreferences/schema-evolution.mdrollback rules when changing the graph
Prerequisites
Graph 3.0 adds the practical contract for new projects; known 2.x project semantics remain readable until explicit owner adoption. Specialists and stage approvals are unchanged. Preliminary policy feasibility is collection before Architecture approval, not early completion of Governance.
Use the graph and apex-recall show <project> --json together. If the returned session is empty, initialize or recover the project within the active owner's scope before routing. The already selected Orchestrator continues explicit new-project requests into Requirements; initialization is not an additional user-selection or approval gate.
Steps
Orchestrator protocol for routing the next step:
- Load
templates/workflow-graph.json - Read current state — use
session.current_step,session.steps, andsession.decisionsin the response. - Resolve the node — use the per-step status map to distinguish Design (
3) from Governance (3_5);
use decisions.iac_tool for the CodeGen and Deploy tracks. Do not derive a node ID from the numeric pointer alone.
- Check node status:
complete→ followon_completeedges → find next nodein_progress→ resume fromsub_stepcheckpointpending→ verify graph prerequisites, human approval and owner readiness before dispatchskipped→ followon_skipedges- failed or blocked → stop and follow the applicable declared recovery or refinement route
- Apply edge and node conditions — honor optional Design selection and the IaC track;
do not execute every outgoing edge when alternatives are present.
- If next is a `gate` — check its preconditions, present to the user, wait for approval, and record the decision.
- If next is a `subagent-fan-out` — use the owning agent's actual role/tool contract;
never infer callable workers from graph children. Collect required outputs before continuing.
- Repeat until all nodes are complete or blocked
Core Concepts
Forward execution and declared refinement returns are distinct. A shared step-4 plan feeds step-5b / step-5t, then step-6b / step-6t. Approval gates remain blocking on both tracks.
Full node-type table, edge-condition matrix, and IaC routing rules in `references/dag-concepts.md`.
Workflow Graph
The full machine-readable DAG is in: templates/workflow-graph.json
Reading the Graph (Orchestrator Protocol)
Use Steps as the routing protocol. Response fields are documented in `show-schema.md`.
Reference Index
VS Code Copilot SDK entrypoints share workflow-entry.md. For parent-to-worker calls, read native worker lifecycle, then execution-subagent.md. Manual Host workflow entry uses apex-workflow-start, including resume [project]. This skill requires explicit owner selection; it does not bind models/tools.
Other operations load only their requested procedure: imported IaC, existing Azure as-built, project review. Published documentation maintenance belongs to the separate jonathan-vella/apex-docs repository. Do not load retired documentation tooling when advancing infrastructure steps. For their shared execution boundary, read operational-safety.md.
| Reference | File | Content | |
|---|---|---|---|
| Workflow Graph | templates/workflow-graph.json | Full DAG for the multi-step workflow | |
| Orchestrator Handoff | references/orchestrator-handoff-guide.md | Gate templates, IaC routing, delegation rules | |
| Subagent Integration | references/subagent-integration.md | Subagent matrix, pricing accuracy, review protocols | |
| Handoff Validation Rules | references/handoff-validation-rules.md | B1a–B5 rule reference (workflow-handoffs PART) | |
| Track Parity Spec | references/track-parity-spec.md | B4 normalization spec for Bicep/Terraform parity | |
| Schema Evolution | references/schema-evolution.md | D1 versioning policy + D2 rollback (metadata.version) |
Validation Surfaces
The workflow graph is enforced at three points:
| Validator | Rule registry | Scope | |
|---|---|---|---|
tools/scripts/validate-workflow-graph.mjs | inline | Graph shape + schema | |
tools/scripts/validate-agents.mjs --only=workflow-handoffs | WORKFLOW_HANDOFF_RULES | Historical UI fixtures and native agents[] dispatch policy | |
tools/scripts/validate-artifacts.mjs | ARTIFACT_HEADINGS["00-handoff.md"] | Gate-companion file H2 sync |
npm run validate:_node includes these checks. For focused graph and handoff checks, use npm run validate:workflow-graph and npm run lint:workflow-handoffs. Artifact Markdown validation remains owned by the existing hooks and challenger review.
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryChecking.github/skills/apex-workflow-engine/SKILL.md
Popularity
- Repository stars
- 216
- Repository forks
- 80
- 7-day trend score
- 32
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/jonathan-vella/apex/apex-workflow-engine)