- Home
- Skills
- Ed3dai
- Ed3d Plugins
- Using Plan And Execute
using-plan-and-execute_skill
- Python
128
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill ed3dai/ed3d-plugins --skill using-plan-and-execute- SKILL.md4.3 KB
Overview
This skill enforces a mandatory plan-and-execute workflow at the start of every conversation. It ensures the agent discovers relevant skills, reads their contents before using them, performs brainstorming before implementation, and turns checklist steps into tracked todos. The goal is to prevent skipped steps and reduce repeated mistakes by making discipline the default behavior.
How this skill works
On first response the agent runs a strict checklist: enumerate available skills, determine applicability, and if a skill applies, announce usage and invoke it with the Skill tool. It requires reading any candidate skill before announcing or executing it. If a skill contains a checklist, the agent must create separate task todos for each checklist item using TaskCreate (or TodoWrite).
When to use it
- Beginning any new user request or task-focused conversation
- Before writing code, tests, or automation for a feature
- When multiple skills or plugins might apply to a request
- Whenever there is potential for skipped procedural steps
- When a task includes verification, debugging, or deployment steps
Best practices
- Always list available skills mentally and check each for applicability before acting
- Read the full skill content before announcing or using it; do not rely on memory
- Announce which skill you will use and why, e.g., "I'm using X to Y"
- For any checklist in a skill, create distinct tracked todos rather than batching or skipping
- Follow rigid skill rules (TDD, verification) exactly; adapt only flexible patterns
Example use cases
- User requests a new feature: enumerate skills, use brainstorming skill, then TDD skill to implement
- Bug report arrives: check for debugging or verification skills, announce and run the debugging workflow
- Infrastructure change: read ops-related skills, create TaskCreate todos for each deployment step
- Quick question that could trigger a skill: still perform the initial skill-check checklist
- Code review request: run the relevant review skill, announce usage, and produce checklist todos
FAQ
If there is even a small chance a skill applies, read it. The rule is conservative: read first, decide second.
Can I skip creating todos for trivial checklist items?
No. Every checklist item must become its own tracked todo to avoid skipped steps and ensure accountability.