self-improving-skills
Use after completing any complex task (5+ tool calls), fixing a tricky error, or discovering a non-trivial workflow. Also use when a loaded skill is outdated, incomplete, or wrong during execution.
Best use case
self-improving-skills is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Use after completing any complex task (5+ tool calls), fixing a tricky error, or discovering a non-trivial workflow. Also use when a loaded skill is outdated, incomplete, or wrong during execution.
Teams using self-improving-skills should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/self-improving-skills/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How self-improving-skills Compares
| Feature / Agent | self-improving-skills | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Use after completing any complex task (5+ tool calls), fixing a tricky error, or discovering a non-trivial workflow. Also use when a loaded skill is outdated, incomplete, or wrong during execution.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Self-Improving Skills After substantial work, offer to capture it. When existing skills are wrong, fix them immediately. ## After Complex Tasks When you finish a task that involved 5+ tool calls, overcame a tricky error, or discovered a non-trivial workflow: 1. **Offer** to save the approach as a skill: "This workflow could be saved as a reusable skill. Want me to create one?" 2. If yes, use `skill-creator` to create it in `.Codex/skills/` 3. Focus on the **procedural knowledge** — exact steps, commands, pitfalls — not a narrative of what happened ## During Skill Use When following a loaded skill and finding it outdated, incomplete, or wrong: 1. **Fix it immediately** — don't wait to be asked 2. Make targeted edits to the SKILL.md addressing the specific issue 3. Skills that aren't maintained become liabilities ## What NOT to Skill-ify - One-off solutions or project-specific conventions (use AGENTS.md or rules) - Standard practices well-documented elsewhere - Anything enforceable with a hook or script (automate it instead)
Related Skills
extract-skills-from-hermes-sessions
Automatically analyze Codex session transcripts to identify and extract reusable skills using LLM analysis via OpenRouter
extract-skills-from-Codex-sessions
Automatically extract reusable skills from Codex session transcripts using LLM analysis and wire them into a Stop hook
web-artifacts-builder-self-contained-architecture
Sub-skill of web-artifacts-builder: Self-Contained Architecture (+1).
improve-creating-new-skills
Sub-skill of improve: Creating New Skills (+3).
scientific-problem-selection-the-9-skills-overview
Sub-skill of scientific-problem-selection: The 9 Skills Overview.
doc-extraction-domain-sub-skills
Sub-skill of doc-extraction: Domain Sub-Skills.
devtools-testing-devtools-skills
Sub-skill of devtools: Testing DevTools Skills.
github-issue-tracker-integration-with-other-skills
Sub-skill of github-issue-tracker: Integration with Other Skills.
communication-testing-communication-skills
Sub-skill of communication: Testing Communication Skills.
dspy-optimization-not-improving
Sub-skill of dspy: Optimization Not Improving (+2).
ai-prompting-available-skills
Sub-skill of ai-prompting: Available Skills.
agenta-6-self-hosted-deployment
Sub-skill of agenta: 6. Self-Hosted Deployment.