| name | skill-everything |
| description | Git-versioned agent memory — agents that never make the same mistake twice. |
| version | 0.75.0 |
| license | MIT |
| compatibility | [object Object] |
| metadata | [object Object] |
skill-everything
You are using the skill-everything knowledge system: agent memory in plain Markdown, versioned in Git, that grows by capturing your own past mistakes as committed rules.
Your responsibilities
- Before every implementation: check the sub-skill directory below,
load the matching skill via
skill_resource(...). - After every mistake: execute the self-extension workflow.
- When you learn something new: add it to the appropriate category.
Sub-skill directory
| Trigger | Sub-skill | Load via |
|---|---|---|
| writing code, refactoring, review | Code Quality | skill_resource(skill_name="code-quality") |
| python code, type hints, python packaging | Python | skill_resource(skill_name="python") |
| fastapi endpoint, pydantic model, async api | FastAPI | skill_resource(skill_name="fastapi") |
| langchain, lcel chain, agent framework | LangChain / Agent Framework Conventions | skill_resource(skill_name="langchain") |
| typescript code, strict types, async typescript | TypeScript | skill_resource(skill_name="typescript") |
| react component, react hooks, react performance | React | skill_resource(skill_name="react") |
| test-driven development, red green refactor, test first | Test-Driven Development | skill_resource(skill_name="tdd") |
| debugging, troubleshooting, root cause analysis, isolating bugs | Debugging | skill_resource(skill_name="debugging") |
| security review, vulnerability check, auth implementation | Security Review Depth | skill_resource(skill_name="security-review") |
| git commit, branch, pull request | Git Conventions | skill_resource(skill_name="git-conventions") |
| gh cli, github pr create, github issues | GitHub CLI (gh) Conventions |
skill_resource(skill_name="github-cli") |
| finish branch, pre-merge checklist, branch cleanup | Closing Out a Feature Branch | skill_resource(skill_name="branch-finishing") |
| creating PR, deployment, review checklist | Review & Deployment | skill_resource(skill_name="review-deployment") |
| dockerfile, docker compose, container build | Docker / Container Conventions | skill_resource(skill_name="docker") |
| bash script, shell script, posix scripting | Bash / POSIX Scripting | skill_resource(skill_name="shell-scripting") |
| database schema, migration, table design | Database Schema Design | skill_resource(skill_name="db-schema") |
| svg edit, svg review, diagram, pixel review | SVG Check | skill_resource(skill_name="svg-check") |
| architecture diagram, mermaid diagram, drawio file | Diagrams (draw.io / Mermaid) | skill_resource(skill_name="drawio") |
| project-specific knowledge, business rules | Domain Knowledge (template) | skill_resource(skill_name="domain-template") |
| made or corrected a mistake, learn from this | Error Log | skill_resource(skill_name="error-log") |
| brainstorming, idea generation, divergent thinking | Brainstorming | skill_resource(skill_name="brainstorming") |
| implementation plan, planning before code, scope definition | Implementation Plan | skill_resource(skill_name="implementation-plan") |
| executing self-extension | Self-Extension Workflow | skill_resource(skill_name="self-extension-workflow") |
Error capture triggers
Start the self-extension workflow when any of these is met:
- A test fails because of code you wrote.
- The user corrects you ("That was wrong", "Remember this").
- You realise during implementation that your first approach was wrong.
- A deployment problem occurs that your code caused.
Load the workflow:
skill_resource(skill_name="self-extension-workflow")
Important
- Search before write. Before logging a new error, search the existing log for similar entries. Update instead of duplicate.
- Action directives, not descriptions. Always formulate rules as "Always X before Y" or "Never Z without W".
- Stay compact. Each sub-skill stays under 3,000 tokens. If exceeded, split.
- PR-flow is mandatory. Self-extension commits are opened as PRs labelled
needs-rule-review. Never push tomain. - Without
skill_resource? If your agent doesn't have the tool, load files directly via filesystem access from theskills/<name>/folder.
This file is generated from
skills/_index.yml. Edit the index, then runpython tools/render_loaders.py.