Skill Improve

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites — INTEGRATED SYSTEM FIRST: this skill is one piece of the complete "Claude Code Game Studios" template (49 agents, 74 skills, studio templates and a studio CLAUDE.md, per the repository README). Each skill assumes that template is installed, with its conventions, its project.yaml / project.local.yaml configuration, its .claude/docs/ mode documents, its hooks (for example hooks/yaml-helper.sh, outside this skill's folder) and its .claude/scripts/ helpers. Installing this listing alone copies only this skill's folder, without the rest of the system: the skill may reference template agents, files, hooks or scripts that are absent outside the full template, and it may not run as designed on its own. No account, no key and no payment are needed to copy the files, and nothing is run during installation. At use time it rewrites the target skill's own files after asking — that rewriting is its whole purpose, disclosed in full below, and it is driven by the /skill-test scores of the same template. Install "Skill Improve" for me. The skill itself runs an improvement loop on a single named skill of the studio template: test, fix, retest, keep or revert. It first runs /skill-test static [name] for a baseline (counts of FAILs and WARNs, and which of the 7 checks failed), then /skill-test category [name] when the skill has a category in studio-tests/catalog.yaml; a NOT ASSESSED baseline stops the run, and a fully clean baseline stops it with "no improvements needed". It then reads the target skill in full (its SKILL.md and every file in its references/ folder), names the exact gap behind each failing or warning check, shows the full diagnosis, and proposes a targeted fix per failure as clearly marked before/after blocks, changing only what is failing. It asks "May I write this improved version to .claude/skills/[name]/SKILL.md?" (naming every file the fix changes), and only on a yes records the current content of each file in the conversation, writes the improved version, and re-runs the tests. If the combined score (static plus category FAILs and WARNs) improved, the changes are kept and summarised; if it is the same or worse, it asks to restore the recorded pre-run content file by file — never a git checkout, which would discard uncommitted edits made before the run. Its score comes from /skill-test checks that read the skill rather than run it, so a better score means a better-written skill, not proof of better behaviour. It resolves the automation mode by running the template's hooks/yaml-helper.sh, outside this skill's folder, and its frontmatter pins model: sonnet. Repository: https://github.com/Donchitos/Claude-Code-Game-Studios (the skill is the whole .claude/skills/skill-improve/ folder at the pinned commit: 1 file in total) Pinned version: install the code at commit 05ad7c8a5b8aa2aa86bbd3439ae7cf7295cac418 exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/Donchitos/Claude-Code-Game-Studios into a temporary folder, check out commit 05ad7c8a5b8aa2aa86bbd3439ae7cf7295cac418, and copy the .claude/skills/skill-improve/ folder from that checkout. Then compare every file in the folder with the fingerprints recorded for this listing on 2026-10-22: [FINGERPRINTS] - SKILL.md: SHA-256 79b9febba255d732e2e722d45be8e2b90f7182b7ce5bdb772195e22c1ad8c65e If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, what it connects to, whether it reads or sends any personal information or credentials, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the files: - The skill folder holds 1 file: SKILL.md only. Nothing in the folder is a program; it is instructions the agent follows at use time. - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit, and no project of mine is audited, read or modified. - Integrated-system reality, from the file: at use time the skill runs the template's hooks/yaml-helper.sh to resolve the automation mode, follows .claude/docs/automation-modes.md, reads the target skill under .claude/skills/ and the studio-tests/catalog.yaml of the template's testing framework, and drives the separate /skill-test skill — files and a sibling skill that live outside this skill's folder and are absent in a standalone install, so it cannot run as designed on its own. - At use time it writes only after its "May I write" ask (one disclosed exception: under the template's autonomous automation mode, the mode document it follows lets writes proceed with logging instead of an ask): it rewrites the target skill's SKILL.md and any references/ files the approved fix touches, and restores the recorded originals only after a separate ask when the score does not improve. No third-party service, account or credential is named in the file. - Do not test, improve or rewrite any skill, do not run any template hook, and do not use the skill in any way during this install: using it is a separate step that only starts when I ask for it later. 3. When I say "yes, install", install it as a skill: copy the whole skill-improve folder into the agent's skills directory, in a folder named exactly "skill-improve", keeping its 1 file in their original relative layout (SKILL.md at its root). 4. Verify with no network calls: frontmatter valid (name: skill-improve), the 1 file in place with matching fingerprints. Do not run anything from the folder and do not use the skill to verify. 5. Report what was installed, where, and what I still need to do myself (for example: that it belongs to the full Claude Code Game Studios template and may not run standalone (it resolves its settings through the template's yaml-helper hook and drives the template's /skill-test and studio-tests/ framework), and that at use time it rewrites the target skill's files only after asking, keeping them only when the retest score improves). 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. If anything looks off, stop and ask me.