---
title: "workflow"
description: "Reusable engineering reviews, audits, handoffs, and maintenance workflows"
canonical: https://agentpluginsdirectory.com/plugins/workflow--appelent
last-updated: 2026-09-18
---

# workflow
Reusable engineering reviews, audits, handoffs, and maintenance workflows
- Slug: workflow--appelent
- Publisher: AppElent
- Repository: https://github.com/AppElent/agent-plugins
- Manifest: plugins/workflow/plugin.json
- Version: 1.0.0
- License: MIT
- Category (editorial): other
- Skills: 11 (cleaning-up-codebases, feature-gap-analysis, handoff-session, model-selector, review-app, review-session, self-improve, the-orca-way, upgrade-deps, usage-audit, workflow)
- MCP servers: 0
- Stars: 0
- Repository created: 2026-09-18
- Repository last pushed: 2026-09-18
- Publisher type: User
- Listing: https://agentpluginsdirectory.com/plugins/workflow--appelent
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What workflow does, in the publisher's words

Reusable engineering workflows that leave no application feature record: reviews, audits, handoffs, model selection, dependency maintenance, and self-improvement.

Claude Code entry point: /workflow:skill. Codex discovers the included skills from their descriptions.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/AppElent/agent-plugins/HEAD/plugins/workflow/README.md

## Skills

- cleaning-up-codebases: Use when reviewing a codebase for cruft, dead code, anti-patterns, scope creep, or architectural drift, an audit that asks "should this exist?" before "how do we improve it?", tiers the findings, and negotiates with the owner before deleting anything.
- feature-gap-analysis: Compare this product's features against named competitors and file one triaged GitHub issue naming the parity gaps, the missing features, and the game-changers. Use when the user asks how the app stacks up against competitors, what features we're missing, where we're behind, what the competition ju…
- handoff-session: Compact this conversation into a handoff brief on disk and start a fresh session that picks the work up immediately. Use when the context window is filling up mid-task, when the user says "hand off", "/handoff-session", "start a new session on this", "continue this in a fresh session", or when the…
- model-selector: Advise which provider, model and reasoning effort to use for a coding task across the Codex and Claude subscriptions. Use when the user asks which model to use, which agent should handle a task, whether to switch providers, what effort level to run at, or says "/model-selector", "which model for th…
- review-app: Autonomously drive a local app through the active provider's browser automation, exercise it end-to-end, and produce a self-contained findings report. Use when asked to review an app, run an end-to-end review, find UI bugs automatically, or test a newly built flow.
- review-session: Start a live review session: I browse/test the app in the preview, you capture my feedback and file a self-contained GitHub issue. Use when I say "let's do a review session", "start reviewing", or similar.
- self-improve: Extract durable lessons from the current session, or sweep the project's past sessions when asked, and route each one to the right knowledge layer (project CLAUDE.md, global CLAUDE.md, auto memory, an existing skill, or a new skill). Use when the user asks to \"self-improve\", \"distill this sessio…
- the-orca-way: Route work through Orca-managed surfaces instead of the agent's private ones, Orca worktrees not raw git worktrees, Orca terminals not in-process subagents or background Bash, Orca's browser and editor not hidden tool panes. Use when, and only when, the user explicitly invokes it, appends "the orca…
- upgrade-deps: Upgrade this app's dependencies safely: check outdated packages, prioritize the TanStack ecosystem, verify typecheck/lint/tests/build, then commit. Use when asked to "upgrade dependencies", "update packages", "bump versions", or similar.
- usage-audit: Audit a codebase for how it behaves against usage-metered service quotas, Convex function calls, database bandwidth and storage, Cloudflare Workers requests and CPU, and rank findings by what they scale with rather than by today's reading. Use when asked to optimize usage, cut service costs, stay i…
- workflow: Front door for reusable engineering workflows. Use when the user asks to list or choose AppElent workflow capabilities, capture a durable workflow lesson, or route work involving reviews, audits, handoffs, dependency maintenance, model selection, or self-improvement.

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.