leamsigc/shortsgenerator/workflow-conventions
Claude Code and agent skill by leamsigc/shortsgenerator
- community
Checking that the SKILL.md file is still in the repository.
Workflow structure, environment variable patterns, planning rules, and development flow for this project. Relevant when creating, editing, or debugging workflows.
Install
npx vskill@latest install leamsigc/shortsgenerator/workflow-conventionsSKILL.md
Workflow Structure
Every workflow directory must follow this layout:
Dependencies are managed in the root requirements.txt, not per-workflow.
Deployment Targets
Every workflow must be runnable locally first. Modal is an optional additional deployment layer. Specify the target in the plan and CATALOG.md.
| Target | How to Run | Notes | |
|---|---|---|---|
| Local | python workflows/{name}/main.py | Always supported — the baseline for every workflow | |
| Local + Modal | Local run + modal deploy workflows/{name}/main.py | Adds Modal deployment on top of local. Use for scheduled, GPU, or API-deployed functions |
Modal Modes
When a workflow targets Modal, it must also declare its Modal mode(s) in the plan and CATALOG.md. A single Modal function cannot be both a web endpoint and a scheduled/manual function — use separate entry points when combining modes.
| Mode | Entry Point | Trigger | |
|---|---|---|---|
| Webhook | run_modal() with @modal.fastapi_endpoint | HTTP POST with bearer token | |
| Scheduled | run_scheduled() with schedule=modal.Cron(...) | Automatic cron + Modal dashboard "Run now" | |
| Manual | run_manual() as plain @app.function | Modal dashboard "Run now" or modal run CLI |
Modes can be combined (e.g., webhook + scheduled). All entry points share the same run() core logic and image.
Environment Variables
Env vars use a two-tier loading system:
- Root `.env` (project root) — loaded first. Shared vars across workflows.
- Workflow `.env` (
workflows/{name}/.env) — loaded second withoverride=True. Workflow-specific overrides.
Planning
- Save plans to
.agent/plans/{workflow-name}.md - Plans are independent — no sequencing, no cross-workflow dependencies
- Each plan must include:
- What the workflow does (inputs, outputs, side effects)
- Dependencies and services needed
- Deployment target (Local, or Local + Modal)
- Validation steps to verify it works
- Complexity indicator:
- Simple - Single-pass executable, low risk
- Medium - May need iteration, some complexity
- Complex - Break into sub-tasks before executing
- Use
.agent/plans/_template.mdas a starting point
Development Flow
- Scaffold — Run
/new-workflowto create the directory and plan template - Plan — Fill in the plan at
.agent/plans/{workflow-name}.md - Build — Run
/buildto execute the plan - Validate — Test the workflow runs correctly
- Ship — Deploy to target (Modal, cron, etc.)
Progress Tracking
See CATALOG.md for the status of all workflows. Updated when a workflow is completed.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/workflow-conventions/SKILL.md
Popularity
- Repository stars
- 348
- Repository forks
- 75
- 7-day trend score
- 29
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/leamsigc/shortsgenerator/workflow-conventions)