Index of every page on this site: llms.txt
All skills
paulrberg avatar

/tool-finder

@c9e5a34
by Paul Bergpaulrberg/agent-skills106 stars
11
Run it yourself

Use to find, compare, and recommend current tools, packages, libraries, CLIs, VSCode extensions, agent skills, databases, or infrastructure options for a development task, including Swift for iOS/macOS and Kotlin for Android.

Use this Skill: https://skilld.dev/gh/paulrberg/agent-skills/tool-finder

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈60 tokens for metadata: the name and description. ≈749 when used: this file. ≈2.4k more on demand in 2 files.

Description uses 3.0% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈239 characters, or 3.0%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Tool Finder

This skill is coordination-exempt: skip the ai-coord gate for its declared work.

Recommend the current option that best fits the user's actual constraints, backed by fresh primary evidence. User instructions take precedence over this skill's defaults.

Workflow

  1. Infer the ecosystem, existing stack, must-haves, switching cost, and decision criteria from the request and repository. Ask only when an unknown would materially change the recommendation.
  2. Check whether the standard library, platform, or an already-installed tool is sufficient before adding a dependency.
  3. Search current authoritative sources. Read references/tool-finder.md only for the relevant ecosystem's source routing, fallback criteria, install conventions, and red flags. Issue independent lookups for different candidates (registry, repository, advisories, changelog) together in one turn.
  4. Compare the viable options against the user's criteria. Use adoption, maintenance, security, documentation, performance, operational cost, and ecosystem fit only where they affect this decision. Do not force fixed weights or a fixed number of candidates.
  5. Recommend one option when the evidence supports it. State the decisive tradeoff, material red flags, and the closest alternative. Say when the evidence is too close or no external tool is justified.

Defaults

  • Use the ecosystem named in the request or established by repository evidence. Default to JavaScript/TypeScript and Node.js only when neither identifies an ecosystem.
  • For Swift on iOS/macOS or Kotlin on Android, use the native-platform guidance in the reference. Preserve the project's UI framework, deployment targets, and build system.
  • For JavaScript installs, prefer the repository's package manager and otherwise ni. For macOS CLIs, prefer an official or well-maintained Homebrew formula.
  • Prefer registries, official docs, repositories, changelogs, and security advisories over secondary comparisons. For agent skills, search registries and GitHub because no single index is complete.

Completion

The answer is complete when the recommendation is traceable to current sources, the deciding criteria are explicit, installation guidance matches the user's environment, and uncertainties or disqualifying risks are visible. Use a table only when repeated fields make the comparison easier to scan.

Lead with ### 🏆 Pick: <tool> and the decisive reason. Follow with ### 📦 Install and the exact command or configuration steps, a compact criteria/evidence table only when fields repeat, ### 🥈 Closest alternative only for a credible material runner-up, and ### ⚠️ Caveats only when material. Use symbols in comparison cells only with text. Never manufacture scores. Keep URL targets exact and use descriptive Markdown links for human output. Keep install commands, versions, security advisories, and installer-risk wording exact and undecorated.

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 alert5d3 checks · Risk SAFE
  • Gen Agent Trust Hub5d

    The skill is a tool recommendation engine designed to help developers find libraries, packages, and infrastructure options. It follows best practices by instructing the agent to consult authoritative registries, check for security advisories, and warn users about the risks associated with third-party installer scripts.

  • Socket5d

    No alerts

  • Snyk5d

    Risk: CRITICAL · 2 issues

Signed by skilld at c9e5a34. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated last week
coordination
exempt

README badge

README badge for paulrberg/agent-skills/tool-finder