Run History Skill Upgrader

A meta-skill that turns real run logs, validation failures and user feedback into a concrete upgrade plan for an existing skill, and only edits files after explicit approval.

Dev & CodingIntermediate★ 701⑂ 58AI score 8/10Last updated: Aug 9, 2026

What it does

A maintenance-only skill for refactoring skills you already have when they keep repeating the same mistakes, reference stale docs, or have bloated into unreadable instruction piles.

Sorts inputs into strong evidence (run outputs, logs, tests, diffs, user corrections) and weak evidence (one-off timeouts, unverified search snippets, model-invented ideas).
Classifies signals: process gap, validation gap, source drift, platform drift, routing gap, content bloat, user preference.
Applies a generalization gate so a single incident does not automatically become a permanent rule.
Recommends one net-improvement shape: no_change, maintenance_note_only, prune_or_consolidate, local_refactor, cross_reference_refactor, major_refactor, or deprecate_or_replace_source.
Always stops at plan stage first. Vague go-aheads like "just do it" do not count; approval must point at the specific plan.
After applying: runs the bundled validator, py_compile/json.tool checks, trigger and route regression, then reports files changed, ideas rejected, and remaining risks.

Who it's for

Builders maintaining a personal or team library of custom skills.
Anyone whose skills misfire, over-trigger, or keep growing without getting better.
Teams that need an approval trail before shared skill files are modified.

Examples

"My deploy skill skipped the same validation step in the last three runs" → classifies the gap and proposes a plan with regression cases.
"The vendor API docs moved to v2 but our skill still assumes v1" → tagged as source drift, proposes deleting/replacing the stale guidance.
"This SKILL.md ballooned to 800 lines" → produces a prune-and-consolidate plan, then applies it after approval and re-verifies trigger behavior.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/dongshuyan/compass-skills/HEAD/skills/run-history-skill-upgrader/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)

Download the ZIP with the button below.
In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.

↓ Download ZIP

Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/run-history-skill-upgrader folder from the GitHub repo dongshuyan/compass-skills into my ~/.claude/skills/run-history-skill-upgrader/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/dongshuyan/compass-skills.git && mkdir -p ~/.claude/skills && cp -r compass-skills/skills/run-history-skill-upgrader ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

Open your terminal (Terminal on macOS/Linux, PowerShell on Windows).
Clone the repo: git clone https://github.com/dongshuyan/compass-skills.git
Create the skills folder: mkdir -p ~/.claude/skills
Copy the skill: cp -r compass-skills/skills/run-history-skill-upgrader ~/.claude/skills/
Verify that references/, scripts/, and evals/ came along — the bundled validator needs them.
Check Python is available: python3 --version
Restart Claude Code and ask something like "improve skill X based on the last few runs".
You will get a plan only. Review it, then approve explicitly with "apply the plan above".

View source on GitHub ↗License: MIT