Self-Improving Agent
Turn failures, corrections, and validated wins into audited behavior changes — capture candidates, run evals, apply lessons deliberately
- What
- Turn failures, corrections, and validated wins into audited behavior changes — capture candidates, run evals, apply lessons deliberately
- Cost
- Free
- Needs
- none beyond a working agent
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — a skill for turning evidence from completed work into small, auditable behavior changes. Use it after a failure, a user correction, a repeated workflow problem, or a validated success reveals a reusable lesson: capture bounded, redacted candidates, run executable behavior evals, and keep validation separate from application. Every run ends in exactly one state — candidate, validated, applied, rejected, superseded, rolled back, no-delta, or open-question. The default result is a candidate or no change, not an automatic rewrite of skills: an applied lesson must change future behavior and come with a representative check proving it. By @zhaono1, listed here with credit to its creator. Honest caveats: a discipline framework, not an automatic learning system — it deliberately resists rewriting durable guidance without evidence and a named owner; not for routine session summaries or speculative ideas without evidence. Skill Harbor never reviews the code, review it yourself before use.
Version:
Install
Prerequisites: none beyond a working agent Install "Self-Improving Agent" for me. Turn failures, corrections, and validated wins into reviewed and validated behavior changes — capture candidates, run evals, apply lessons deliberately Repository: https://github.com/zhaono1/agent-playbook/blob/main/skills/self-improving-agent/SKILL.md 1. Fetch the SKILL.md file for the zhaono1-self-improving-agent skill from the repository into a temporary folder and summarize what it does in one or two sentences. 2. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). This repo should contain zero secrets in code, credentials only via the secure vault, allowed hosts declared in the SKILL.md. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "zhaono1-self-improving-agent". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. any accounts, API keys, or CLI tools mentioned in the prerequisites above). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. Never ask me to paste secrets in chat — credentials go through the secure vault or environment variables. If anything looks off, stop and ask me.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.