[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7BSDA3CNHHP07666X3H90CB":3,"skill-same:01M7BSDA3CNHHP07666X3H90CB":86},{"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":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7BSDA3CNHHP07666X3H90CB","shell-script-expert",null,"Use when the user asks \"review this script\", \"write a bash script\", \"fix my shell script\", \"make this script safer\", \"why does this script fail\", \"convert this to a script\", or working with .sh\u002F.bash files, sh\u002Fbash syntax, shellcheck, cronjobs, and bash one-liners.","thimslugga\u002Fagent-plugins","thimslugga","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F285055?v=4","01M7BRYXXABGR31PBN8YW9YYD3","Adam Kaminski","01M7BRYXY1GHRY2VCK15WG8NCZ",0,"MIT","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: shell-script-expert\ndescription: >-\n  Use when the user asks \"review this script\", \"write a bash script\", \"fix my\n  shell script\", \"make this script safer\", \"why does this script fail\",\n  \"convert this to a script\", or working with .sh\u002F.bash files, sh\u002Fbash syntax,\n  shellcheck, cronjobs, and bash one-liners.\n---\n\n# Shell Script Expert\n\nShell is glue. It starts processes, routes their output, and checks whether they worked. Every strength follows from \nthat, and most failures come from asking it to do more.\n\nThree rules govern the whole skill:\n\n- **Quote every expansion.** Unquoted variables are the single largest source\n  of shell bugs.\n- **Write Google style.** The\n  [Google Shell Style Guide](https:\u002F\u002Fgoogle.github.io\u002Fstyleguide\u002Fshellguide.html)\n  is the house standard: `#!\u002Fbin\u002Fbash`, 2 space indent, 80 columns,\n  `\"${braced}\"` variables, `main \"$@\"` last. The rules that change generated\n  code are in `references\u002Fgoogle-style.md`; read it before writing.\n- **Ship shellcheck-clean.** A script is not finished until\n  `scripts\u002Fcheck.sh` passes. Reading the code is not a substitute for running\n  the linter.\n\n## Step 1: Confirm shell is the right tool\n\nStay in shell when the work is invoking other programs, wiring pipes, testing\nexit codes, and moving files. Package installs, service restarts, log\nrotation, backup wrappers, CI glue, and container entrypoints all belong here.\n\nLeave shell when any of these appear:\n\n| Signal                                                            | Better tool         |\n| ----------------------------------------------------------------- | ------------------- |\n| Past 100 lines, or control flow that is no longer straightforward | Python              |\n| Parsing JSON, YAML, XML, or CSV beyond a single `jq` call         | Python              |\n| Arithmetic beyond counters, or any floating point                 | Python              |\n| Data structures: nested maps, records, sorting by field           | Python              |\n| Retry with backoff plus structured logging plus state             | Python              |\n| Performance matters                                               | Python, Go, or Rust |\n| Needs to run where no interpreter is installed                    | Go or Rust          |\n\nThe 100 line threshold comes from the style guide, and the reasoning behind it\nmatters more than the number: scripts grow, and rewriting early is far cheaper\nthan rewriting later. State the recommendation before writing the script, not\nafter.\n\n## Step 2: Set the portability target\n\n- **`#!\u002Fbin\u002Fbash`** is the default. Put shell options on a `set` line rather\n  than in the shebang, so that `bash script.sh` behaves the same way.\n- **`#!\u002Fusr\u002Fbin\u002Fenv bash`** only when the fleet forces it, most often because\n  macOS is in scope and `\u002Fbin\u002Fbash` there is version 3.2.\n- **`#!\u002Fbin\u002Fsh`** only when POSIX portability is a stated requirement. It\n  costs arrays, `[[`, `local`, and `pipefail`. Do not pay that price by\n  accident.\n\nMinimal container images strip more than expected, so guard external tools\nwith `command -v` rather than assuming they exist. `references\u002Fportability.md`\ncovers bash version gates, GNU versus BSD tool differences, and Amazon Linux\n2023 base images.\n\n## Step 3: Build from the template\n\nCopy `assets\u002Ftemplate.sh` as the starting skeleton. It supplies the file\nheader, strict mode, `usage`, stderr logging, `require_cmds`, `trap`-based\ncleanup, dry-run support, Google-style function header comments, and argument\nparsing that already handles `--` and unknown options correctly.\n\nUse `assets\u002Fboilerplate.sh` as a pattern catalogue, not as a file to copy\nwhole. It intentionally contains independent examples that should be selected\nand trimmed to keep the resulting script small.\n\nStrip what the script does not need. An unused `--verbose` flag is noise, not\nsafety.\n\n### Rules that hold in every script\n\n**Streams have jobs.** Data goes to stdout, errors and progress go to stderr.\nThis keeps the script usable inside a pipeline and keeps cron from mailing\nroutine chatter.\n\n**Exit codes are an interface.** 0 for success, 1 for general failure, 2 for\nusage error. Leave 126 and 127 to the shell, and preserve `128 + signal` when\nhandling signals.\n\n**Strict mode is a backstop, not error handling.** Bash scripts use\n`set -Eeuo pipefail`; POSIX `sh` scripts use `set -eu` because `-E` and\n`pipefail` are not portable. These modes silently decline to fire in several\ncommon situations. Read `references\u002Fstrict-mode.md` before relying on them,\nparticularly before writing `local var=$(command)`, piping into `head`, or\nreading `$?` after `if ! command`.\n\n**Respect blast radius.** For anything that deletes, overwrites, restarts, or\nsends: validate the target path, use `rm -rf -- \"${target:?target is unset}\"`\nso an empty variable aborts instead of erasing the filesystem, expand\nwildcards as `.\u002F*` rather than `*`, and put destructive tools behind a\n`--dry-run` flag. When the script is risky and the user has not asked for a\nflag, add one anyway and say why.\n\n**Treat input as hostile.** Anything from a form, a header, a filename, an\nenvironment variable, or a CI variable can carry shell metacharacters. Pass\nvalues as arguments rather than assembling command strings, and validate\nagainst an allowlist. See `references\u002Fsecurity.md`.\n\n**Temp files come from `mktemp`.** Never `\u002Ftmp\u002Fmyscript.$$`. Register cleanup\nwith `trap` in the same breath as the creation.\n\n**Loop over `find -print0`, never over `ls`.** Filenames contain spaces and\nnewlines. Piping into `while` also loses every variable the loop sets, so use\nprocess substitution or `readarray` instead.\n\n**Use `printf`, not `echo -e`.** `echo` flag handling varies between shells\nand builds; `printf '%s\\n'` does not.\n\n## Step 4: Verify\n\nRun the checker on every script produced or modified:\n\n```bash\n.\u002Fscripts\u002Fcheck.sh path\u002Fto\u002Fscript.sh\n```\n\nIt runs `bash -n` for syntax, `shellcheck` for correctness, and\n`shfmt -i 2 -ci -bn -d` for Google-style formatting when those are installed,\nand prints install commands when they are not. Fix every finding, or add a\ntargeted `# shellcheck disable=SCxxxx` with a comment explaining the reason on\nthe line above. Blanket disables at the top of a file hide real bugs. Look up\nany unfamiliar code at\n\u003Chttps:\u002F\u002Fgithub.com\u002Fkoalaman\u002Fshellcheck\u002Fwiki\u002F>, one page per rule.\n\nThen exercise the failure paths, not only the happy path: missing argument,\nmissing input file, unwritable output directory, and a command that exits\nnon-zero in the middle. Most shell bugs live there.\n\n## Reviewing or debugging an existing script\n\nRun `scripts\u002Fcheck.sh` first. ShellCheck finds most of what a manual read\nwould find, faster and without ego, so the conversation can start with the\nfindings rather than with style opinions.\n\nThen work through `references\u002Freview-checklist.md`, which orders the audit by\nseverity: destructive operations, then injection and quoting, then error\nhandling, then style and portability. Report findings in that order and lead\nwith anything that could destroy data.\n\nFor a script that misbehaves rather than fails, `set -x` narrows it fastest.\n`PS4='+ ${BASH_SOURCE}:${LINENO}: '` makes the trace readable, and\n`bash -x script.sh` avoids editing the file at all.\n\n## Additional resources\n\n- **`assets\u002Ftemplate.sh`**: production skeleton, copy and trim\n- **`assets\u002Fboilerplate.sh`**: extended pattern catalogue, select and trim\n- **`scripts\u002Fcheck.sh`**: syntax, lint, and format gate\n- **`references\u002Fgoogle-style.md`**: the house style, and where this skill\n  extends it\n- **`references\u002Fstrict-mode.md`**: where `set -Eeuo pipefail` fails to fire\n- **`references\u002Fsecurity.md`**: command injection, environment tampering,\n  secrets, privilege\n- **`references\u002Fpatterns.md`**: argument parsing, locking, retries,\n  parallelism, safe temp files, config handling, JSON with jq\n- **`references\u002Fportability.md`**: bash version gates, GNU versus BSD tools,\n  minimal container images\n- **`references\u002Freview-checklist.md`**: severity-ordered audit for existing\n  scripts\n\nExternal sources this skill is built on:\n\n- Google Shell Style Guide:\n  \u003Chttps:\u002F\u002Fgoogle.github.io\u002Fstyleguide\u002Fshellguide.html>\n- ShellCheck: \u003Chttps:\u002F\u002Fwww.shellcheck.net\u002F>\n- ShellCheck rule explanations: \u003Chttps:\u002F\u002Fgithub.com\u002Fkoalaman\u002Fshellcheck\u002Fwiki\u002F>\n- Bash FAQ, Greg's Wiki: \u003Chttp:\u002F\u002Fmywiki.wooledge.org\u002FBashFAQ>\n- OWASP command injection:\n  \u003Chttps:\u002F\u002Fowasp.org\u002Fwww-community\u002Fattacks\u002FCommand_Injection>\n\nSuggest a rewrite in Python whenever the script starts reaching for these\nreferences to work around a limitation rather than to use a feature.\n","# Shell Script Expert\n\nShell is glue. It starts processes, routes their output, and checks whether they worked. Every strength follows from \nthat, and most failures come from asking it to do more.\n\nThree rules govern the whole skill:\n\n- **Quote every expansion.** Unquoted variables are the single largest source\n  of shell bugs.\n- **Write Google style.** The\n  [Google Shell Style Guide](https:\u002F\u002Fgoogle.github.io\u002Fstyleguide\u002Fshellguide.html)\n  is the house standard: `#!\u002Fbin\u002Fbash`, 2 space indent, 80 columns,\n  `\"${braced}\"` variables, `main \"$@\"` last. The rules that change generated\n  code are in `references\u002Fgoogle-style.md`; read it before writing.\n- **Ship shellcheck-clean.** A script is not finished until\n  `scripts\u002Fcheck.sh` passes. Reading the code is not a substitute for running\n  the linter.\n\n## Step 1: Confirm shell is the right tool\n\nStay in shell when the work is invoking other programs, wiring pipes, testing\nexit codes, and moving files. Package installs, service restarts, log\nrotation, backup wrappers, CI glue, and container entrypoints all belong here.\n\nLeave shell when any of these appear:\n\n| Signal                                                            | Better tool         |\n| ----------------------------------------------------------------- | ------------------- |\n| Past 100 lines, or control flow that is no longer straightforward | Python              |\n| Parsing JSON, YAML, XML, or CSV beyond a single `jq` call         | Python              |\n| Arithmetic beyond counters, or any floating point                 | Python              |\n| Data structures: nested maps, records, sorting by field           | Python              |\n| Retry with backoff plus structured logging plus state             | Python              |\n| Performance matters                                               | Python, Go, or Rust |\n| Needs to run where no interpreter is installed                    | Go or Rust          |\n\nThe 100 line threshold comes from the style guide, and the reasoning behind it\nmatters more than the number: scripts grow, and rewriting early is far cheaper\nthan rewriting later. State the recommendation before writing the script, not\nafter.\n\n## Step 2: Set the portability target\n\n- **`#!\u002Fbin\u002Fbash`** is the default. Put shell options on a `set` line rather\n  than in the shebang, so that `bash script.sh` behaves the same way.\n- **`#!\u002Fusr\u002Fbin\u002Fenv bash`** only when the fleet forces it, most often because\n  macOS is in scope and `\u002Fbin\u002Fbash` there is version 3.2.\n- **`#!\u002Fbin\u002Fsh`** only when POSIX portability is a stated requirement. It\n  costs arrays, `[[`, `local`, and `pipefail`. Do not pay that price by\n  accident.\n\nMinimal container images strip more than expected, so guard external tools\nwith `command -v` rather than assuming they exist. `references\u002Fportability.md`\ncovers bash version gates, GNU versus BSD tool differences, and Amazon Linux\n2023 base images.\n\n## Step 3: Build from the template\n\nCopy `assets\u002Ftemplate.sh` as the starting skeleton. It supplies the file\nheader, strict mode, `usage`, stderr logging, `require_cmds`, `trap`-based\ncleanup, dry-run support, Google-style function header comments, and argument\nparsing that already handles `--` and unknown options correctly.\n\nUse `assets\u002Fboilerplate.sh` as a pattern catalogue, not as a file to copy\nwhole. It intentionally contains independent examples that should be selected\nand trimmed to keep the resulting script small.\n\nStrip what the script does not need. An unused `--verbose` flag is noise, not\nsafety.\n\n### Rules that hold in every script\n\n**Streams have jobs.** Data goes to stdout, errors and progress go to stderr.\nThis keeps the script usable inside a pipeline and keeps cron from mailing\nroutine chatter.\n\n**Exit codes are an interface.** 0 for success, 1 for general failure, 2 for\nusage error. Leave 126 and 127 to the shell, and preserve `128 + signal` when\nhandling signals.\n\n**Strict mode is a backstop, not error handling.** Bash scripts use\n`set -Eeuo pipefail`; POSIX `sh` scripts use `set -eu` because `-E` and\n`pipefail` are not portable. These modes silently decline to fire in several\ncommon situations. Read `references\u002Fstrict-mode.md` before relying on them,\nparticularly before writing `local var=$(command)`, piping into `head`, or\nreading `$?` after `if ! command`.\n\n**Respect blast radius.** For anything that deletes, overwrites, restarts, or\nsends: validate the target path, use `rm -rf -- \"${target:?target is unset}\"`\nso an empty variable aborts instead of erasing the filesystem, expand\nwildcards as `.\u002F*` rather than `*`, and put destructive tools behind a\n`--dry-run` flag. When the script is risky and the user has not asked for a\nflag, add one anyway and say why.\n\n**Treat input as hostile.** Anything from a form, a header, a filename, an\nenvironment variable, or a CI variable can carry shell metacharacters. Pass\nvalues as arguments rather than assembling command strings, and validate\nagainst an allowlist. See `references\u002Fsecurity.md`.\n\n**Temp files come from `mktemp`.** Never `\u002Ftmp\u002Fmyscript.$$`. Register cleanup\nwith `trap` in the same breath as the creation.\n\n**Loop over `find -print0`, never over `ls`.** Filenames contain spaces and\nnewlines. Piping into `while` also loses every variable the loop sets, so use\nprocess substitution or `readarray` instead.\n\n**Use `printf`, not `echo -e`.** `echo` flag handling varies between shells\nand builds; `printf '%s\\n'` does not.\n\n## Step 4: Verify\n\nRun the checker on every script produced or modified:\n\n```bash\n.\u002Fscripts\u002Fcheck.sh path\u002Fto\u002Fscript.sh\n```\n\nIt runs `bash -n` for syntax, `shellcheck` for correctness, and\n`shfmt -i 2 -ci -bn -d` for Google-style formatting when those are installed,\nand prints install commands when they are not. Fix every finding, or add a\ntargeted `# shellcheck disable=SCxxxx` with a comment explaining the reason on\nthe line above. Blanket disables at the top of a file hide real bugs. Look up\nany unfamiliar code at\n\u003Chttps:\u002F\u002Fgithub.com\u002Fkoalaman\u002Fshellcheck\u002Fwiki\u002F>, one page per rule.\n\nThen exercise the failure paths, not only the happy path: missing argument,\nmissing input file, unwritable output directory, and a command that exits\nnon-zero in the middle. Most shell bugs live there.\n\n## Reviewing or debugging an existing script\n\nRun `scripts\u002Fcheck.sh` first. ShellCheck finds most of what a manual read\nwould find, faster and without ego, so the conversation can start with the\nfindings rather than with style opinions.\n\nThen work through `references\u002Freview-checklist.md`, which orders the audit by\nseverity: destructive operations, then injection and quoting, then error\nhandling, then style and portability. Report findings in that order and lead\nwith anything that could destroy data.\n\nFor a script that misbehaves rather than fails, `set -x` narrows it fastest.\n`PS4='+ ${BASH_SOURCE}:${LINENO}: '` makes the trace readable, and\n`bash -x script.sh` avoids editing the file at all.\n\n## Additional resources\n\n- **`assets\u002Ftemplate.sh`**: production skeleton, copy and trim\n- **`assets\u002Fboilerplate.sh`**: extended pattern catalogue, select and trim\n- **`scripts\u002Fcheck.sh`**: syntax, lint, and format gate\n- **`references\u002Fgoogle-style.md`**: the house style, and where this skill\n  extends it\n- **`references\u002Fstrict-mode.md`**: where `set -Eeuo pipefail` fails to fire\n- **`references\u002Fsecurity.md`**: command injection, environment tampering,\n  secrets, privilege\n- **`references\u002Fpatterns.md`**: argument parsing, locking, retries,\n  parallelism, safe temp files, config handling, JSON with jq\n- **`references\u002Fportability.md`**: bash version gates, GNU versus BSD tools,\n  minimal container images\n- **`references\u002Freview-checklist.md`**: severity-ordered audit for existing\n  scripts\n\nExternal sources this skill is built on:\n\n- Google Shell Style Guide:\n  \u003Chttps:\u002F\u002Fgoogle.github.io\u002Fstyleguide\u002Fshellguide.html>\n- ShellCheck: \u003Chttps:\u002F\u002Fwww.shellcheck.net\u002F>\n- ShellCheck rule explanations: \u003Chttps:\u002F\u002Fgithub.com\u002Fkoalaman\u002Fshellcheck\u002Fwiki\u002F>\n- Bash FAQ, Greg's Wiki: \u003Chttp:\u002F\u002Fmywiki.wooledge.org\u002FBashFAQ>\n- OWASP command injection:\n  \u003Chttps:\u002F\u002Fowasp.org\u002Fwww-community\u002Fattacks\u002FCommand_Injection>\n\nSuggest a rewrite in Python whenever the script starts reaching for these\nreferences to work around a limitation rather than to use a feature.\n",{"name":5,"description":7},"main","62177b0f31651c18aefb738bccb54168b8a5a982","skills\u002Fshell-script-expert","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins","https:\u002F\u002Fgithub.com\u002Fthimslugga","https:\u002F\u002Fadamkaminski.com","I am interested in Linux, macOS, Python, Golang, Rust and Networking.\r\n\r\nOpinions expressed are my own and may not be the views of my employer.","2026-11-11 02:10:30",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add thimslugga\u002Fagent-plugins --skill shell-script-expert",{"tool":38,"command":39},"bunx","bunx skills add thimslugga\u002Fagent-plugins --skill shell-script-expert",{"tool":41,"command":42},"pnpm","pnpm dlx skills add thimslugga\u002Fagent-plugins --skill shell-script-expert","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002FSKILL.md",true,[46,52,56,62,66,70,74,78,82],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"assets\u002Fboilerplate.sh","SCRIPT","脚本",29744,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Fassets\u002Fboilerplate.sh",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"assets\u002Ftemplate.sh",5498,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Fassets\u002Ftemplate.sh",{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Fgoogle-style.md","REFERENCE","参考文档",6235,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Freferences\u002Fgoogle-style.md",{"path":63,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Fpatterns.md",7882,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Freferences\u002Fpatterns.md",{"path":67,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Fportability.md",6719,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Freferences\u002Fportability.md",{"path":71,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":72,"sourceHtmlUrl":73},"references\u002Freview-checklist.md",6352,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Freferences\u002Freview-checklist.md",{"path":75,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":76,"sourceHtmlUrl":77},"references\u002Fsecurity.md",4908,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Freferences\u002Fsecurity.md",{"path":79,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":80,"sourceHtmlUrl":81},"references\u002Fstrict-mode.md",5276,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Freferences\u002Fstrict-mode.md",{"path":83,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":84,"sourceHtmlUrl":85},"scripts\u002Fcheck.sh",6873,"https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fshell-script-expert\u002Fscripts\u002Fcheck.sh",[87,94,101,107,114,121],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":19},"01M7BSD6STC24Q13N4J8E0T93X","prose-synthesize","Turn unstructured notes into a structured, actionable plan. Use when given brain dumps, stream-of-consciousness, or scattered thoughts needing order.","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fprose-synthesize\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":19},"01M7BSD16KRRWFMM3EN1WPK8G3","prose-distill","Condense verbose text to its essence.  Use when asked to condense, tighten, shorten, reduce verbosity, or omit needless words while preserving substance.","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fprose-distill\u002FSKILL.md",[],[],{"skillNo":102,"name":103,"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":104,"categories":105,"keywords":106,"publishedAt":19},"01M7BSCYJWSP5VJK3822XRNXR2","mindcraft","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fmindcraft\u002FSKILL.md",[],[],{"skillNo":108,"name":109,"nameZh":6,"description":110,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":111,"categories":112,"keywords":113,"publishedAt":19},"01M7BSCRQ1GCDCTPNZ3WA3R0JE","learn","Tutor for intellectual understanding: teach how or why something works rather than complete a task or provide an evaluative verdict.\nTrigger for explicit learning requests (teach, explain, ELI5, walkthroughs, quizzes, flashcards, definitions); terse concept prompts; confusion signals; prerequisite or sequencing questions; and conceptual questions about mechanisms, causes, or dynamics.\nDo not trigger for task completion (coding, writing, calculation, or translation), factual or news lookup, personal troubleshooting, resource recommendations, or opinion and interpretive verdicts.","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Flearn\u002FSKILL.md",[],[],{"skillNo":115,"name":116,"nameZh":6,"description":117,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":118,"categories":119,"keywords":120,"publishedAt":19},"01M7BSCNAY02WX6BGDW06Y60D9","deployment-readiness","Use to check deployment readiness of a host.","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fskills\u002Fdeployment-readiness\u002FSKILL.md",[],[],{"skillNo":122,"name":123,"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":124,"categories":125,"keywords":126,"publishedAt":19},"01M7BSCJFYYTJP7146V4MHY3KV","david-8","https:\u002F\u002Fgithub.com\u002Fthimslugga\u002Fagent-plugins\u002Fblob\u002Fmain\u002Fplugins\u002Fweyland-yutani\u002Fskills\u002Fdavid-8\u002FSKILL.md",[],[]]