[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DHW74TCQ7RX3Y0CV5JQP54":3,"skill-same:01M6DHW74TCQ7RX3Y0CV5JQP54":110},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":16,"sourceHtmlUrl":17,"categories":18,"keywords":24,"publishedAt":48,"linkOnly":49,"markdownContent":50,"markdownBody":51,"frontMatter":52,"refName":59,"commitSha":60,"skillDirectory":61,"forksCount":62,"repositoryHtmlUrl":63,"repositoryHomepageUrl":64,"ownerHtmlUrl":65,"ownerBlogUrl":66,"ownerDescription":67,"updatedAt":68,"install":69,"resources":82},"01M6DHW74TCQ7RX3Y0CV5JQP54","skill-install-improved",null,"Install an improved variant of one named skill: resolve it by local path, repo, or name, run skill-creator's retrofit on a throwaway copy, then install the improved result. Don't use for improving in place, upstream PRs, or plain asm install.","luongnv89\u002Fasm","luongnv89","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F3288457?v=4","01M6DHVDRYW79PQVAJZBW4V18N","Luong NGUYEN","01M6DHVDYJK5E6GDHG3EQ0NCBH",975,"MIT","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002FSKILL.md",[19],{"categoryNo":20,"categoryName":21,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47],"skill install","install","skill install improved","upstream","result","throwaway","prs","skill-creator","variant","for","path","name","local","run","use","edit","write","read","repo","grep","Bash","git","code","2026-11-10 10:12:30",false,"---\nname: skill-install-improved\ndescription: \"Install an improved variant of one named skill: resolve it by local path, repo, or name, run skill-creator's retrofit on a throwaway copy, then install the improved result. Don't use for improving in place, upstream PRs, or plain asm install.\"\nlicense: MIT\ncompatibility: \"Claude Code; requires `asm` and `git` on PATH\"\nallowed-tools: Bash Read Write Edit Grep Glob\neffort: high\ndependencies:\n  - skill-creator\nmetadata:\n  version: 1.3.0\n  author: luongnv89\n---\n\n# Skill Install Improved\n\nYou install an **improved** variant of one skill instead of the skill as published: resolve the target to a local directory → run `skill-creator`'s Subpath B1 retrofit loop on it → install the improved result → report what was improved and from what. The improvement is a step inside the install, not a separate errand. One skill per run.\n\n## When to Use\n\n- The user says \"install `\u003Cskill>` but improve it first\" or \"install an improved version of `\u003Cskill>`\"\n- The user points at a catalog skill below the 85\u002F8 floor and wants the better version on their machine\n- The user has a local or cloned skill directory and wants the improved variant installed, not the raw one\n\nDo **not** trigger for: improving without installing (`skill-creator`), contributing upstream (`skill-upstream-pr`), authoring from scratch (`skill-creator`), plain installs (`asm install \u003Csource>`), or bulk-improving many skills at once.\n\n## Prerequisites\n\nVerify each before resolving anything. If one fails, stop and tell the user which one.\n\n- `asm` and `git` on PATH, and Python 3 (skill-creator's validator needs it)\n- Network access to GitHub, for the repo and name forms\n- Write access to the install target for the chosen tool and scope (see `asm list --json` for existing install paths)\n\n### The user's own files are never modified (design decision)\n\n**Every input form is improved on a throwaway copy under `$(mktemp -d)` — local paths included.** An install that rewrites the user's working copy is a side effect they did not ask for, and the retrofit loop's `git pull --rebase` would rebase their branch.\n\n**Repo sync is skipped for every form.** A local-path copy has no `origin`. A repo or name clone does have `origin`, but it is a throwaway `--depth 1` checkout that may sit on a detached `--ref`, and a pull could move it off the requested ref. The retrofit loop allows this skip for a caller's throwaway copy. Say so when you skip it.\n\n**Trade-off:** the improvement lands only in the installed copy. To persist it, point the user at `skill-creator` (their own path) or `skill-upstream-pr` (the source repo).\n\n## Dependency Preflight (mandatory)\n\nThis skill invokes `skill-creator`, declared in frontmatter `dependencies`. Before Phase 0, verify `asm` is available:\n\n```bash\ncommand -v asm >\u002Fdev\u002Fnull || {\n  echo \"Missing installer: npm install -g agent-skill-manager\" >&2\n  exit 1\n}\nasm deps discover skill-install-improved --json\n```\n\nWhen execution reaches Phase 1, the main agent runs:\n\n```bash\nasm deps acquire skill-creator --session \u003Ccaller-session-id> --json\n```\n\n`\u003Ccaller-session-id>` is a unique id the main agent picks once per run, e.g. `skill-install-improved-\u003Cepoch-seconds>`; reuse it for the release.\n\nSet `SKILL_CREATOR_DIR` to the directory holding the returned `skillMdPath`, and use it immediately; do not wait for a provider catalog rescan. The retrofit loop is `$SKILL_CREATOR_DIR\u002Freferences\u002Fretrofit-loop.md`; its validator is `$SKILL_CREATOR_DIR\u002Fscripts\u002Fquick_validate.py`. If the acquire fails, stop before Phase 1 and report the error — this skill cannot improve anything without it.\n\nThe main agent releases the lease from its own `finally`\u002Fshutdown handling — after Phase 4, or at any earlier stop:\n\n```bash\nasm deps release --session \u003Ccaller-session-id> --json\n```\n\nIf Phase 0 stops before Phase 1, do not acquire.\n\n## Inputs\n\nThe user identifies **one** target skill, in any of four forms — a local path (`skills\u002Ffoo`), a local `SKILL.md` file path, a repo (`https:\u002F\u002Fgithub.com\u002Fowner\u002Frepo`, `owner\u002Frepo`), or a skill name (`code-review`). Phase 0 normalizes all of them to one local directory, `$SKILL_PATH`.\n\nAlso needed before Phase 3: the install **scope** (`global` or `project`) and the target **tool** (`-p\u002F--tool` — `claude`, `codex`, `agents`, …). If the user did not state one, ask; never guess. `asm install` hard-fails without `--tool` in a non-interactive run, and `-y` does not cover it.\n\nOptional: a `--name \u003Calt>` intent.\n\nBefore Phase 0, record the directory the user invoked you from as `CALLER_DIR` (`pwd`). Project-scope paths resolve from the current directory, so Phase 3 must run there, not inside `$WORK`.\n\n## Workflow\n\nExecute phases in order. Do not skip or reorder.\n\n### Phase 0 — Resolve the target to one local directory\n\nFull contract in `references\u002Ftarget-resolution.md`. In short, with `WORK=\"$(mktemp -d)\"`:\n\n- **Local path** — `cp -R` into `$WORK`. A `SKILL.md` **file** path folds to its parent first; the retrofit loop takes a directory, not a file.\n- **Repo** — plain `git clone` into `$WORK`. **Never `gh repo fork`** — this skill performs no public GitHub action.\n- **Skill name** — `asm search \"\u003Cterm>\" --available --json`, then copy the string after `asm install ` from the chosen result's `installCommand` **verbatim**. Never hand-construct `github:owner\u002Frepo:path`. Clone what it names.\n\nSet `$SKILL_PATH` to the directory holding the chosen `SKILL.md`. If the checkout holds several and no subpath was given, list them (`find \"$WORK\" -maxdepth 5 -name SKILL.md -type f`) and **ask which one. Never guess.**\n\nRecord for the report: the supplied identifier, the resolved install or clone URL, and the upstream commit SHA.\n\n### Phase 1 — Delegate to skill-creator's retrofit loop\n\nThis skill does not reimplement the improvement loop, and it reads that loop only now, so the loop's references stay out of the context budget until Phase 1. Acquire `skill-creator` (Dependency Preflight), then follow `$SKILL_CREATOR_DIR\u002Freferences\u002Fretrofit-loop.md` (Subpath B1) with `$SKILL_PATH` as the target, through its Phase 7 report.\n\nTwo adaptations, because the target is a throwaway copy:\n\n- Skip its \"Repo Sync Before Edits\" step. Log `— repo sync skipped (throwaway copy)` and continue.\n- `.asm-improver\u002F` is written **relative to the current working directory**. Run the loop with cwd inside `$SKILL_PATH`, or Phase 2 has nothing to harvest.\n- If the loop asks how to handle a target with no frontmatter, answer \"abort\" (see Edge Cases). If it offers a Subpath B3 delegation conversion, decline: this run installs, it does not restructure.\n\nIf the baseline already clears both gates, the retrofit loop stops without editing — a valid outcome, see Edge Cases.\n\n### Phase 2 — Harvest artifacts, then clean up\n\n**Harvest before any cleanup.** Read the files under `$SKILL_PATH\u002F.asm-improver\u002F` that `references\u002Finstall-and-report.md` lists, and extract its fields. An early-exit baseline writes no `iter-N.json`; then the baseline is also the final state.\n\nThen confirm `$SKILL_PATH` is still inside `$WORK`, and remove the two artifacts that must not ship (`asm install` copies the source recursively):\n\n```bash\nrm -f \"$SKILL_PATH\u002FSKILL.md.bak\"\nrm -rf \"$SKILL_PATH\u002F.asm-improver\" \"$SKILL_PATH\u002F.git\"\n```\n\nNever point either command at a user-supplied path.\n\n### Phase 3 — Install the improved directory\n\nThe retrofit loop never renames, so the improved variant keeps the original frontmatter `name`. **`asm install` never refuses on a collision**: with `-y` it deletes and replaces the target directory without a prompt. So probe **before** invoking it:\n\nRun this phase from `CALLER_DIR`:\n\n```bash\ncd \"$CALLER_DIR\"\nasm list --json    # is this skill's name or directory already installed for $TOOL \u002F $SCOPE?\n```\n\nRead `references\u002Finstall-and-report.md` → _Collision policy_ and take the path it gives for the probe result. The rule that never changes: if the probe shows any match — frontmatter `name` or directory name — name the target path and what occupies it, and get **explicit confirmation before running `asm install`**. Use `--name \u003Calt>` only when the user asks for it, after the duplicate-trigger warning.\n\nOnly after the probe is read and any collision is confirmed:\n\n```bash\nasm install \"$SKILL_PATH\" -p \"$TOOL\" --scope \"$SCOPE\" --json -y\n```\n\nInstall `$SKILL_PATH`, never the original source. Take the installed path from the command's `--json` output (`.path`); never assume `~\u002F.claude\u002Fskills\u002F`.\n\n### Phase 4 — Report, then clean up\n\nFill in the report template in `references\u002Finstall-and-report.md`. It covers the four items the user needs: the result, the evidence, what stays untested, and the next decision.\n\nRemove `$WORK` only after the report is printed, then release the dependency lease.\n\n## Step Completion Reports (mandatory)\n\nAfter each phase, print a compact status block: `√` pass, `×` fail, `—` context.\n\n```\n◆ Phase N — [phase name] ([N of 5])\n··································································\n  [check 1]:         √ pass\n  [check 2]:         × fail — [reason]\n  Result:            PASS | FAIL | PARTIAL\n```\n\nChecks per phase:\n\n- **Phase 0** — `Form identified`, `Copied to temp`, `SKILL_PATH unambiguous`, `Provenance recorded`\n- **Phase 1** — `skill-creator acquired`, `Baseline captured`, `Repo sync skipped`, `Gates cleared or stop reason known`\n- **Phase 2** — `Metrics harvested`, `.bak removed`, `.asm-improver removed`, `Deletions confined to temp`\n- **Phase 3** — `Collision probed before install`, `Overwrite confirmed (if any)`, `Tool and scope supplied`, `Install succeeded`\n- **Phase 4** — `Report printed`, `Temp dir removed`, `Lease released`\n\n## Acceptance Criteria\n\nThe expected output of a successful run, each item checkable:\n\n- Exactly one target resolved; `$SKILL_PATH` is a directory with a `SKILL.md`, under `$WORK`; the user's files are unmodified\n- Name resolution copied `installCommand` verbatim; repo resolution used plain `git clone` — no fork, no push\n- The retrofit loop ran from the acquired `skill-creator`; `baseline.json`, `iter-N.json`, and `report.md` were read before cleanup\n- The installed skill contains no `.asm-improver\u002F` and no `SKILL.md.bak`\n- `asm install` targeted `$SKILL_PATH` with both `-p\u002F--tool` and `--scope`, after a collision probe; any overwrite was confirmed first\n- The report states the installed path, provenance, and before → after numbers\n- `$WORK` was removed after the report, and the dependency lease was released\n\nWhen reviewing a run's output (by eval or by a human), also check that a reader can:\n\n- **Find the main result** — the first line says an improved variant was installed, the original was installed unchanged, or the run stopped.\n- **Separate facts from assumptions** — measured scores are distinct from the untested runtime behavior of the improved skill.\n- **Trace claims** — before → after numbers map to `baseline.json` and the last `iter-N.json`; the install path maps to install `--json` `.path`.\n- **See the next decision** — the report names any action left (restart the agent, persist the change upstream) or states that none is needed.\n\nAn agent's own review cannot confirm human understanding. A reviewer who received no human feedback records understanding as unconfirmed.\n\n## Edge Cases\n\n- **Baseline already passes both gates** — the retrofit loop stops without editing, but it has already written `.asm-improver\u002F`. Still run Phase 2's cleanup and Phase 3's probe. Install the original, unchanged, and report that no improvement was needed, naming the baseline score. Never imply a delta that did not happen.\n- **Retrofit loop ends in BLOCKER** — show the blocker list and ask whether to install the partial result or abort. Never install one silently.\n- **`skill-creator` cannot be acquired** — stop before Phase 1, remove `$WORK`, and report the `asm deps acquire` error.\n- **Several `SKILL.md` files in a clone** — list them and ask. Never batch, never guess.\n- **`asm search` returns nothing, or several equally-plausible matches** — show the candidates and ask; never install the first hit.\n- **Target has no frontmatter** — `asm eval --fix` cannot add it, and this skill does not scaffold one. Answer \"abort\" to the retrofit loop, report, and stop; do not install an unimprovable skill as if it were improved.\n- **Collision with a same-named skill from another source** — the install would replace someone else's skill with no prompt. Name the target directory and what the probe shows there (`name`, `dirName`, `provider`, `scope`), confirm, and offer `--name \u003Calt>` — all before `asm install` runs.\n- **Clone or disk failure mid-resolve** — stop, remove `$WORK`, report. Never install a partial checkout.\n\n## References\n\n- `references\u002Ftarget-resolution.md` — the input forms normalized to one local `$SKILL_PATH`\n- `references\u002Finstall-and-report.md` — harvested fields, install flags, collision policy, report template\n- `$SKILL_CREATOR_DIR\u002Freferences\u002Fretrofit-loop.md` — the improvement loop this skill delegates to (acquired at run time)\n- `asm install --help` and `asm search --help` — flag references\n","# Skill Install Improved\n\nYou install an **improved** variant of one skill instead of the skill as published: resolve the target to a local directory → run `skill-creator`'s Subpath B1 retrofit loop on it → install the improved result → report what was improved and from what. The improvement is a step inside the install, not a separate errand. One skill per run.\n\n## When to Use\n\n- The user says \"install `\u003Cskill>` but improve it first\" or \"install an improved version of `\u003Cskill>`\"\n- The user points at a catalog skill below the 85\u002F8 floor and wants the better version on their machine\n- The user has a local or cloned skill directory and wants the improved variant installed, not the raw one\n\nDo **not** trigger for: improving without installing (`skill-creator`), contributing upstream (`skill-upstream-pr`), authoring from scratch (`skill-creator`), plain installs (`asm install \u003Csource>`), or bulk-improving many skills at once.\n\n## Prerequisites\n\nVerify each before resolving anything. If one fails, stop and tell the user which one.\n\n- `asm` and `git` on PATH, and Python 3 (skill-creator's validator needs it)\n- Network access to GitHub, for the repo and name forms\n- Write access to the install target for the chosen tool and scope (see `asm list --json` for existing install paths)\n\n### The user's own files are never modified (design decision)\n\n**Every input form is improved on a throwaway copy under `$(mktemp -d)` — local paths included.** An install that rewrites the user's working copy is a side effect they did not ask for, and the retrofit loop's `git pull --rebase` would rebase their branch.\n\n**Repo sync is skipped for every form.** A local-path copy has no `origin`. A repo or name clone does have `origin`, but it is a throwaway `--depth 1` checkout that may sit on a detached `--ref`, and a pull could move it off the requested ref. The retrofit loop allows this skip for a caller's throwaway copy. Say so when you skip it.\n\n**Trade-off:** the improvement lands only in the installed copy. To persist it, point the user at `skill-creator` (their own path) or `skill-upstream-pr` (the source repo).\n\n## Dependency Preflight (mandatory)\n\nThis skill invokes `skill-creator`, declared in frontmatter `dependencies`. Before Phase 0, verify `asm` is available:\n\n```bash\ncommand -v asm >\u002Fdev\u002Fnull || {\n  echo \"Missing installer: npm install -g agent-skill-manager\" >&2\n  exit 1\n}\nasm deps discover skill-install-improved --json\n```\n\nWhen execution reaches Phase 1, the main agent runs:\n\n```bash\nasm deps acquire skill-creator --session \u003Ccaller-session-id> --json\n```\n\n`\u003Ccaller-session-id>` is a unique id the main agent picks once per run, e.g. `skill-install-improved-\u003Cepoch-seconds>`; reuse it for the release.\n\nSet `SKILL_CREATOR_DIR` to the directory holding the returned `skillMdPath`, and use it immediately; do not wait for a provider catalog rescan. The retrofit loop is `$SKILL_CREATOR_DIR\u002Freferences\u002Fretrofit-loop.md`; its validator is `$SKILL_CREATOR_DIR\u002Fscripts\u002Fquick_validate.py`. If the acquire fails, stop before Phase 1 and report the error — this skill cannot improve anything without it.\n\nThe main agent releases the lease from its own `finally`\u002Fshutdown handling — after Phase 4, or at any earlier stop:\n\n```bash\nasm deps release --session \u003Ccaller-session-id> --json\n```\n\nIf Phase 0 stops before Phase 1, do not acquire.\n\n## Inputs\n\nThe user identifies **one** target skill, in any of four forms — a local path (`skills\u002Ffoo`), a local `SKILL.md` file path, a repo (`https:\u002F\u002Fgithub.com\u002Fowner\u002Frepo`, `owner\u002Frepo`), or a skill name (`code-review`). Phase 0 normalizes all of them to one local directory, `$SKILL_PATH`.\n\nAlso needed before Phase 3: the install **scope** (`global` or `project`) and the target **tool** (`-p\u002F--tool` — `claude`, `codex`, `agents`, …). If the user did not state one, ask; never guess. `asm install` hard-fails without `--tool` in a non-interactive run, and `-y` does not cover it.\n\nOptional: a `--name \u003Calt>` intent.\n\nBefore Phase 0, record the directory the user invoked you from as `CALLER_DIR` (`pwd`). Project-scope paths resolve from the current directory, so Phase 3 must run there, not inside `$WORK`.\n\n## Workflow\n\nExecute phases in order. Do not skip or reorder.\n\n### Phase 0 — Resolve the target to one local directory\n\nFull contract in `references\u002Ftarget-resolution.md`. In short, with `WORK=\"$(mktemp -d)\"`:\n\n- **Local path** — `cp -R` into `$WORK`. A `SKILL.md` **file** path folds to its parent first; the retrofit loop takes a directory, not a file.\n- **Repo** — plain `git clone` into `$WORK`. **Never `gh repo fork`** — this skill performs no public GitHub action.\n- **Skill name** — `asm search \"\u003Cterm>\" --available --json`, then copy the string after `asm install ` from the chosen result's `installCommand` **verbatim**. Never hand-construct `github:owner\u002Frepo:path`. Clone what it names.\n\nSet `$SKILL_PATH` to the directory holding the chosen `SKILL.md`. If the checkout holds several and no subpath was given, list them (`find \"$WORK\" -maxdepth 5 -name SKILL.md -type f`) and **ask which one. Never guess.**\n\nRecord for the report: the supplied identifier, the resolved install or clone URL, and the upstream commit SHA.\n\n### Phase 1 — Delegate to skill-creator's retrofit loop\n\nThis skill does not reimplement the improvement loop, and it reads that loop only now, so the loop's references stay out of the context budget until Phase 1. Acquire `skill-creator` (Dependency Preflight), then follow `$SKILL_CREATOR_DIR\u002Freferences\u002Fretrofit-loop.md` (Subpath B1) with `$SKILL_PATH` as the target, through its Phase 7 report.\n\nTwo adaptations, because the target is a throwaway copy:\n\n- Skip its \"Repo Sync Before Edits\" step. Log `— repo sync skipped (throwaway copy)` and continue.\n- `.asm-improver\u002F` is written **relative to the current working directory**. Run the loop with cwd inside `$SKILL_PATH`, or Phase 2 has nothing to harvest.\n- If the loop asks how to handle a target with no frontmatter, answer \"abort\" (see Edge Cases). If it offers a Subpath B3 delegation conversion, decline: this run installs, it does not restructure.\n\nIf the baseline already clears both gates, the retrofit loop stops without editing — a valid outcome, see Edge Cases.\n\n### Phase 2 — Harvest artifacts, then clean up\n\n**Harvest before any cleanup.** Read the files under `$SKILL_PATH\u002F.asm-improver\u002F` that `references\u002Finstall-and-report.md` lists, and extract its fields. An early-exit baseline writes no `iter-N.json`; then the baseline is also the final state.\n\nThen confirm `$SKILL_PATH` is still inside `$WORK`, and remove the two artifacts that must not ship (`asm install` copies the source recursively):\n\n```bash\nrm -f \"$SKILL_PATH\u002FSKILL.md.bak\"\nrm -rf \"$SKILL_PATH\u002F.asm-improver\" \"$SKILL_PATH\u002F.git\"\n```\n\nNever point either command at a user-supplied path.\n\n### Phase 3 — Install the improved directory\n\nThe retrofit loop never renames, so the improved variant keeps the original frontmatter `name`. **`asm install` never refuses on a collision**: with `-y` it deletes and replaces the target directory without a prompt. So probe **before** invoking it:\n\nRun this phase from `CALLER_DIR`:\n\n```bash\ncd \"$CALLER_DIR\"\nasm list --json    # is this skill's name or directory already installed for $TOOL \u002F $SCOPE?\n```\n\nRead `references\u002Finstall-and-report.md` → _Collision policy_ and take the path it gives for the probe result. The rule that never changes: if the probe shows any match — frontmatter `name` or directory name — name the target path and what occupies it, and get **explicit confirmation before running `asm install`**. Use `--name \u003Calt>` only when the user asks for it, after the duplicate-trigger warning.\n\nOnly after the probe is read and any collision is confirmed:\n\n```bash\nasm install \"$SKILL_PATH\" -p \"$TOOL\" --scope \"$SCOPE\" --json -y\n```\n\nInstall `$SKILL_PATH`, never the original source. Take the installed path from the command's `--json` output (`.path`); never assume `~\u002F.claude\u002Fskills\u002F`.\n\n### Phase 4 — Report, then clean up\n\nFill in the report template in `references\u002Finstall-and-report.md`. It covers the four items the user needs: the result, the evidence, what stays untested, and the next decision.\n\nRemove `$WORK` only after the report is printed, then release the dependency lease.\n\n## Step Completion Reports (mandatory)\n\nAfter each phase, print a compact status block: `√` pass, `×` fail, `—` context.\n\n```\n◆ Phase N — [phase name] ([N of 5])\n··································································\n  [check 1]:         √ pass\n  [check 2]:         × fail — [reason]\n  Result:            PASS | FAIL | PARTIAL\n```\n\nChecks per phase:\n\n- **Phase 0** — `Form identified`, `Copied to temp`, `SKILL_PATH unambiguous`, `Provenance recorded`\n- **Phase 1** — `skill-creator acquired`, `Baseline captured`, `Repo sync skipped`, `Gates cleared or stop reason known`\n- **Phase 2** — `Metrics harvested`, `.bak removed`, `.asm-improver removed`, `Deletions confined to temp`\n- **Phase 3** — `Collision probed before install`, `Overwrite confirmed (if any)`, `Tool and scope supplied`, `Install succeeded`\n- **Phase 4** — `Report printed`, `Temp dir removed`, `Lease released`\n\n## Acceptance Criteria\n\nThe expected output of a successful run, each item checkable:\n\n- Exactly one target resolved; `$SKILL_PATH` is a directory with a `SKILL.md`, under `$WORK`; the user's files are unmodified\n- Name resolution copied `installCommand` verbatim; repo resolution used plain `git clone` — no fork, no push\n- The retrofit loop ran from the acquired `skill-creator`; `baseline.json`, `iter-N.json`, and `report.md` were read before cleanup\n- The installed skill contains no `.asm-improver\u002F` and no `SKILL.md.bak`\n- `asm install` targeted `$SKILL_PATH` with both `-p\u002F--tool` and `--scope`, after a collision probe; any overwrite was confirmed first\n- The report states the installed path, provenance, and before → after numbers\n- `$WORK` was removed after the report, and the dependency lease was released\n\nWhen reviewing a run's output (by eval or by a human), also check that a reader can:\n\n- **Find the main result** — the first line says an improved variant was installed, the original was installed unchanged, or the run stopped.\n- **Separate facts from assumptions** — measured scores are distinct from the untested runtime behavior of the improved skill.\n- **Trace claims** — before → after numbers map to `baseline.json` and the last `iter-N.json`; the install path maps to install `--json` `.path`.\n- **See the next decision** — the report names any action left (restart the agent, persist the change upstream) or states that none is needed.\n\nAn agent's own review cannot confirm human understanding. A reviewer who received no human feedback records understanding as unconfirmed.\n\n## Edge Cases\n\n- **Baseline already passes both gates** — the retrofit loop stops without editing, but it has already written `.asm-improver\u002F`. Still run Phase 2's cleanup and Phase 3's probe. Install the original, unchanged, and report that no improvement was needed, naming the baseline score. Never imply a delta that did not happen.\n- **Retrofit loop ends in BLOCKER** — show the blocker list and ask whether to install the partial result or abort. Never install one silently.\n- **`skill-creator` cannot be acquired** — stop before Phase 1, remove `$WORK`, and report the `asm deps acquire` error.\n- **Several `SKILL.md` files in a clone** — list them and ask. Never batch, never guess.\n- **`asm search` returns nothing, or several equally-plausible matches** — show the candidates and ask; never install the first hit.\n- **Target has no frontmatter** — `asm eval --fix` cannot add it, and this skill does not scaffold one. Answer \"abort\" to the retrofit loop, report, and stop; do not install an unimprovable skill as if it were improved.\n- **Collision with a same-named skill from another source** — the install would replace someone else's skill with no prompt. Name the target directory and what the probe shows there (`name`, `dirName`, `provider`, `scope`), confirm, and offer `--name \u003Calt>` — all before `asm install` runs.\n- **Clone or disk failure mid-resolve** — stop, remove `$WORK`, report. Never install a partial checkout.\n\n## References\n\n- `references\u002Ftarget-resolution.md` — the input forms normalized to one local `$SKILL_PATH`\n- `references\u002Finstall-and-report.md` — harvested fields, install flags, collision policy, report template\n- `$SKILL_CREATOR_DIR\u002Freferences\u002Fretrofit-loop.md` — the improvement loop this skill delegates to (acquired at run time)\n- `asm install --help` and `asm search --help` — flag references\n",{"name":5,"effort":53,"license":15,"metadata":54,"description":7,"dependencies":56,"allowed-tools":57,"compatibility":58},"high",{"author":9,"version":55},"1.3.0",[32],"Bash Read Write Edit Grep Glob","Claude Code; requires `asm` and `git` on PATH","main","65fbfb16db2df92eeabe0d8aa36a3ca9c97f81cc","skills\u002Fskill-install-improved",87,"https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm","https:\u002F\u002Fluongnv.com\u002Fasm","https:\u002F\u002Fgithub.com\u002Fluongnv89","https:\u002F\u002Fluongnv.com","Agentic AI & Security Engineer, Paris. Tools and guides for AI coding agents: claude-howto, asm, IDD. Learn, build, share, connect.","2026-11-14 21:37:38",{"source":8,"skillName":5,"commands":70,"prompt":80,"zipAvailable":81},[71,74,77],{"tool":72,"command":73},"npx","npx skills add luongnv89\u002Fasm --skill skill-install-improved",{"tool":75,"command":76},"bunx","bunx skills add luongnv89\u002Fasm --skill skill-install-improved",{"tool":78,"command":79},"pnpm","pnpm dlx skills add luongnv89\u002Fasm --skill skill-install-improved","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002FSKILL.md",true,[83,89,93,98,102,106],{"path":84,"resourceType":85,"resourceTypeDesc":86,"sizeBytes":87,"sourceHtmlUrl":88},"docs\u002FREADME.md","REFERENCE","参考文档",5942,"https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002Fdocs\u002FREADME.md",{"path":90,"resourceType":85,"resourceTypeDesc":86,"sizeBytes":91,"sourceHtmlUrl":92},"evals\u002Fevals.json",3490,"https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002Fevals\u002Fevals.json",{"path":94,"resourceType":95,"resourceTypeDesc":96,"sizeBytes":97,"sourceHtmlUrl":16},"LICENSE.txt","LICENSE","许可证",1066,{"path":99,"resourceType":85,"resourceTypeDesc":86,"sizeBytes":100,"sourceHtmlUrl":101},"references\u002Fdependency_analysis.md",5775,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fscience-research-data-analysis\u002Flatest-release-community\u002Fdepmap\u002Freferences\u002Fdependency_analysis.md",{"path":103,"resourceType":85,"resourceTypeDesc":86,"sizeBytes":104,"sourceHtmlUrl":105},"references\u002Finstall-and-report.md",10183,"https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002Freferences\u002Finstall-and-report.md",{"path":107,"resourceType":85,"resourceTypeDesc":86,"sizeBytes":108,"sourceHtmlUrl":109},"references\u002Ftarget-resolution.md",6068,"https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fskills\u002Fskill-install-improved\u002Freferences\u002Ftarget-resolution.md",[111,130,139,161,181,190],{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":48},"01M6DHXQFRH26RCJBCFGCWKCF2","well-formed","Review pull request diffs for code smells, style issues, and safety problems before merging.","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fwell-formed\u002FSKILL.md",[],[118,42,119,34,120,121,122,123,124,44,125,126,47,127,128,129],"well formed","style","request","fixtures","review pull request","review","code smells","diffs","pull request","issues","pull","Safety",{"skillNo":131,"name":132,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":133,"categories":134,"keywords":135,"publishedAt":48},"01M6DHXN16BAWWPQZM7RQ6GPKC","missing-frontmatter","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fmissing-frontmatter\u002FSKILL.md",[],[136,137,138],"missing","missing frontmatter","frontmatter",{"skillNo":140,"name":141,"nameZh":6,"description":142,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":143,"categories":144,"keywords":148,"publishedAt":48},"01M6WQBQGDN448GBBCPHFZQQAZ","hello-world","A minimal test skill that greets the user and demonstrates the ASM publish workflow.","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fhello-world\u002FSKILL.md",[145],{"categoryNo":146,"categoryName":147,"taxonomyType":22,"taxonomyTypeDesc":23},"01JBSEEDCAT000000000000008","测试与质量",[149,150,151,152,153,154,47,155,156,157,158,159,160],"hello world","hello","world","workflow","test","Test Skill","demo","publish","user","minimal","test, demo","minimal test skill",{"skillNo":162,"name":163,"nameZh":6,"description":164,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":165,"categories":166,"keywords":172,"publishedAt":48},"01M6DHXJV89CFQT467DJXHYG3Z","runtime-budget-pass","Short description that names a clear trigger. Don't use for unrelated docs.","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fsrc\u002Feval\u002Fproviders\u002Fskill-best-practice\u002Fv1\u002Ffixtures\u002Fruntime-budget-pass\u002FSKILL.md",[167],{"categoryNo":168,"categoryName":169,"taxonomyType":170,"taxonomyTypeDesc":171},"01JBSEEDCAT000000000000040","财务","PROFESSION","适用职业",[173,174,175,176,177,153,178,39,179,180,34],"runtime budget pass","runtime","Budget","unrelated","docs","short","clear","trigger",{"skillNo":182,"name":183,"nameZh":6,"description":184,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":185,"categories":186,"keywords":188,"publishedAt":48},"01M6DHXGJ8Q1C5W2WGY47K8MGJ","runtime-budget-fail","AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fsrc\u002Feval\u002Fproviders\u002Fskill-best-practice\u002Fv1\u002Ffixtures\u002Fruntime-budget-fail\u002FSKILL.md",[187],{"categoryNo":168,"categoryName":169,"taxonomyType":170,"taxonomyTypeDesc":171},[174,189,175,153],"runtime budget fail",{"skillNo":191,"name":192,"nameZh":6,"description":193,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":194,"categories":195,"keywords":196,"publishedAt":48},"01M6DHXDJ2SJ6R8AJPXCW6GMN8","metadata-version-semver-pass","Validate a skill. Don't use for unrelated docs.","https:\u002F\u002Fgithub.com\u002Fluongnv89\u002Fasm\u002Fblob\u002Fmain\u002Fsrc\u002Feval\u002Fproviders\u002Fskill-best-practice\u002Fv1\u002Ffixtures\u002Fmetadata-version-semver-pass\u002FSKILL.md",[],[197,198,199,200,153,201,39,176,34,177],"metadata version semver pass","version","metadata","semver","validate"]