Workflow Reference logo

Workflow Reference

Community
neuromechanist
workflow-reference

This skill should be used when executing the epic-dev workflow, creating epic branches, managing sprint phases, working with git worktrees for phased feature development, or when the user mentions "epic dev", "epic-dev", "/epic-dev", "epic workflow", "sprint phases", "phased development", or "git worktree workflow".

Overview

Publisherneuromechanist
Repositoryresearch-skills
Skill nameworkflow-reference
Stars
51
Forks
9
Bundled files
Instructions only
LicenseBSD-3-Clause
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by neuromechanist on GitHub. Read the source before you install it.

Installation

Install the Workflow Reference AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/neuromechanist/research-skills/tree/main/plugins/project/skills/workflow-reference
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/neuromechanist/research-skills.git /tmp/research-skills
mkdir -p .claude/skills
cp -r /tmp/research-skills/plugins/project/skills/workflow-reference .claude/skills/workflow-reference
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Workflow Reference in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Workflow Reference on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Workflow Reference is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Epic/Sprint Development Workflow Reference

Procedural reference for multi-phase feature development using git worktrees, GitHub issues with sub-issues, and phased PR delivery.

Phase Sizing and Epic-vs-Single Decision

One phase = one independently reviewable and testable PR: roughly one subsystem, one migration, or up to ~500 net lines of non-generated change.

  • The description contains 2+ such units -> epic (this workflow).
  • Exactly one unit -> single-phase shortcut (below), or simply the engineering-loop skill without epic machinery.
  • Unsure how to break a description into phases: each phase must have its own testable Definition of Done; if two candidate phases can only be tested together, they are one phase.

Slug derivation for names: take the phase title, drop stopwords, keep 2-3 words with the object noun first, kebab-case ("backend-metrics", "sync-engine"; never "implement-the-new-backend-metrics").

Related skills, invoked by their fully qualified project:<skill> names through the host's skill-invocation mechanism: implementation-planning (writing each phase plan), agent-fanout (parallel implementers, one worktree each), engineering-loop (the inner loop within a phase), debugging (when a phase's tests fail for unclear reasons).

Model handoff by phase

Follow agent-fanout driver mode. The session driver (Claude Sonnet at xhigh, Codex Luna at max) orchestrates the workflow mechanically. Epic decomposition, architecture, and high-risk or contested findings go to Opus (or Fable) or Sol at xhigh. Named Sonnet or Luna workers implement the approved brief in the isolated worktree; fresh Sonnet or Luna reviewers review each phase per the review-convergence rules (two rounds at most). Close/remove one-off agents after their reports are incorporated.

Branch Strategy Decision Tree

Determine the integration branch and branching model:

  1. Integration branch detection:

    • If develop branch exists (local or remote): integration branch = develop
    • Otherwise: integration branch = default branch (main or master)
  2. Epic branch:

    • Named: feature/issue-{N}-epic-{short-name}
    • Created from: integration branch
    • Purpose: collects all phase PRs before final merge to integration
  3. Phase branches:

    • Named: feature/issue-{N}-phase{X}-{short-name}
    • Created from: epic branch
    • Merged to: epic branch (squash merge)
  4. Single-phase shortcut:

    • If only one phase, skip the epic branch layer
    • Create feature branch directly from integration branch
    • Merge directly to integration branch

State File Format

Persistent state stored in .claude/epic.local.md (gitignored via .claude/*.local.md):

yaml
---
epic_issue: 132
epic_title: "Feature: Community Dashboard"
integration_branch: develop
epic_branch: feature/issue-132-epic-dashboard
worktree_base: "../epic-dashboard"
phases:
  - number: 1
    title: "Backend metrics collection"
    issue: 133
    branch: "feature/issue-133-phase1-metrics"
    status: complete       # pending | in_progress | complete
    pr: 135
  - number: 2
    title: "Dashboard frontend"
    issue: 134
    branch: "feature/issue-134-phase2-frontend"
    status: pending
    pr: null
current_phase: 2
created_at: "2026-02-02T12:00:00Z"
---

## Notes
Running notes about the epic, decisions made, blockers encountered.

Status transitions: pending -> in_progress -> complete

Git Worktree Operations

All worktree paths use absolute paths (Bash calls do not persist cd).

Create epic worktree:

bash
REPO_ROOT=$(git rev-parse --show-toplevel)
PARENT=$(dirname "$REPO_ROOT")
git worktree add "$PARENT/epic-{short-name}" -b feature/issue-{N}-epic-{short-name} {integration_branch}

Create phase worktree from epic branch:

bash
git worktree add "$PARENT/{short-name}-phase{X}" -b feature/issue-{N}-phase{X}-{short-name} feature/issue-{EPIC}-epic-{epic-name}

Clean up worktree after merge:

bash
git worktree remove "$PARENT/{worktree-name}"
git branch -d feature/issue-{N}-phase{X}-{short-name}

Handle existing worktree: Before creating, check:

bash
git worktree list | grep -q "{branch-name}" && echo "EXISTS" || echo "NEW"

GitHub Operations

Semantic line breaks remain the default for prose source. Text posted to GitHub (issue and pull-request bodies, comments, reviews, and discussions) is the exception: keep each paragraph and list item on one source line, separate paragraphs with blank lines, and do not insert sentence- or clause-level newlines inside a paragraph.

Create epic issue:

bash
gh issue create --title "Epic: {description}" --label "feature" --body "{body with phase breakdown}"

Create phase sub-issue and link:

bash
PHASE_ISSUE=$(gh issue create --title "Phase {X}: {title}" --label "feature" --body "Part of #{epic_issue}" | grep -o '[0-9]*$')
gh sub-issue add {epic_issue} --sub-issue-number $PHASE_ISSUE

Create PR to epic branch:

bash
gh pr create --base {epic_branch} --title "Phase {X}: {title}" --body "Closes #{phase_issue}\n\nPart of epic #{epic_issue}"

Squash merge (phase PRs into the epic branch only):

bash
gh pr merge --squash --delete-branch

Merge-strategy rule: phase PRs squash-merge into the epic branch to keep it readable; the FINAL epic-to-integration merge uses a regular merge (gh pr merge --merge) to preserve history, unless the user directs otherwise. When a local convention conflicts with the user's standing merge preference, surface the conflict once, before the first merge runs, not after several.

Create final PR (epic to integration):

bash
gh pr create --base {integration_branch} --title "{epic_title}" --body "Closes #{epic_issue}\n\n## Phases completed\n{list}"

Confirmation Points

Confirm (pause and ask):

  • Epic plan: before creating issues and branches, present the full plan
  • Implementation plan: after plan mode (EnterPlanMode) produces the plan, wait for approval
  • Unfixable Critical finding: an accepted Critical finding that cannot be fixed within the phase
  • Final epic merge: before merging epic branch to integration

Auto-proceed (no confirmation):

  • Repository detection
  • Branch and worktree creation (after plan confirmed)
  • Issue and sub-issue creation (after plan confirmed)
  • State file updates
  • Running tests
  • Review fixes, the delta review, and phase merges (per pr-review-toolkit review-convergence.md reference)
  • Recovery from CI failures, conflicts, and silent agents (per engineering-loop unblock-rules.md reference)
  • Worktree cleanup after merge
  • PR creation

Variation Handling

No develop branch

Auto-detected. Use main (or master) as integration branch. All other workflow steps are identical.

Single-phase feature

When the user describes only one phase or says "just one phase":

  • Skip epic worktree creation
  • Create feature branch directly from integration branch: feature/issue-{N}-{short-name}
  • Create single issue (no sub-issues needed)
  • PR targets integration branch directly

Resume interrupted epic

When invoked with --resume:

  • Read .claude/epic.local.md
  • Find the first phase with status != complete
  • Check if its worktree exists; if so, switch to it
  • If worktree does not exist, create it
  • Continue from where the phase left off

Next phase

When invoked with --next-phase:

  • Read state, find next pending phase
  • Mark it in_progress, create worktree, begin execution

Finalize

When invoked with --finalize:

  • Skip to Phase 3 (final PR from epic branch to integration)

Integration with Other Skills/Commands

  • EnterPlanMode: Invoke at the start of each phase for implementation planning (Codex and Copilot: write the plan before editing)
  • project:pr-review-toolkit: Invoke before merging each phase PR and the final epic PR; its review-convergence.md reference caps review at two rounds
  • project:engineering-loop: Optionally follow within a phase for the single-PR change workflow (user decides)
  • TodoWrite: Track all phases and steps throughout

Naming Conventions

Keep names short and lowercase with hyphens:

  • Epic: epic-{2-3 word slug} (e.g., epic-dashboard-metrics)
  • Phase: phase{N}-{2-3 word slug} (e.g., phase1-backend-metrics)
  • Worktree dirs: same as branch slug without feature/issue-N- prefix

Frequently asked questions

What does the Workflow Reference AI skill do?

This skill should be used when executing the epic-dev workflow, creating epic branches, managing sprint phases, working with git worktrees for phased feature development, or when the user mentions "epic dev", "epic-dev", "/epic-dev", "epic workflow", "sprint phases", "phased development", or "git worktree workflow".

Why use Workflow Reference on TypingMind?

Because you install it once and use it with any model. Workflow Reference is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Workflow Reference in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/neuromechanist/research-skills/tree/main/plugins/project/skills/workflow-reference. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Workflow Reference?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Workflow Reference?

As many as you like. As long as a model supports skills, you can use Workflow Reference with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Workflow Reference AI skill free?

Yes. It is published on GitHub by neuromechanist under the BSD-3-Clause license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇