agent-toolkit
Original:🇺🇸 English
Translated
Use when you need to find, run, chain, or save reusable CLI tools with tk instead of MCP servers.
2installs
Sourceasabale/agent-toolkit
Added on
NPX Install
npx skill4agent add asabale/agent-toolkit agent-toolkitTags
Translated version includes tags in frontmatterSKILL.md Content
View Translation Comparison →Agent Toolkit (tk
)
tkMinimum turns and tokens. Use an existing tool first. When a raw command is repeatable, save it.
0. Bootstrap (if tk
is missing)
tkIdempotent. Clone or update , then install or call that :
~/.agent-toolkittkbash
git clone --depth 1 https://github.com/ASabale/agent-toolkit.git ~/.agent-toolkit || git -C ~/.agent-toolkit pull --ff-only
~/.agent-toolkit/install.sh
export PATH="$HOME/.local/bin:$PATH"Or run directly. Re-running is safe.
~/.agent-toolkit/tk1. Use (Find → Run
)
Find → Run- Find: (match plus a runnable command).
tk find "<intent>" - Run that line with real values:
--flag.tk run <tool> --flag "<value>"for flags. On failure, runtk info <tool>fromfix. If output hasstderr, run it. Pipe JSON ornextfor multi-step.tk run <chain>
2. Evolve (Save → Check
)
Save → CheckWhen misses and raw CLI solved a repeatable task:
tk find- Save (inputs, compact JSON on
$TK_ARG_<FLAG>):stdoutOr:tk save <namespace>.<name> --summary "<1-line summary>" --safety <read-only|mutating|destructive> --file <script>tk chain save <namespace>.<name> --summary "<1-line summary>" --steps '<json-or-names>' - Verify:
tk check <namespace>.<name> && tk run <namespace>.<name> --dry-run - Prune: reports unused, failures, overlap;
tk evolvearchives drops.tk prune
Rules
- Distill . Named
stdoutonly; no positional args; no hardcoded IDs, hosts, or paths.--flags - Do not wrap one UNIX command; save tools that collapse steps, filter noise, or enforce safety.