agentleFS
Sign inStart free
D-RoboticsD-Robotics/rdk-skills/skills/rdk-skill-finder/SKILL.md

rdk-skill-finder

Skill5 starsChanged 3 months agoApache-2.0View on GitHub

Find the real D-Robotics RDK Skill or workspace pack for a task, board platform, or installation type, then return its deterministic install or handoff action. Use for requests such as "which skill handles X5 model quantization?", "find an RDK skill for diagnostics", "what RDK toolchain pack should I use?", or "帮我找 X5 模型量化/编译对应的 Skill". Do not use to install or modify anything directly; use rdk-pack-installer for confirmed workspace-pack installation.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Search the bundled, generated RDK catalog and return only registered Skills.

    line 20
  2. Distinctive:

    Use when the user needs to discover the right RDK Skill or workspace pack from a task description, board/platform term, or exact name.

    line 26
  3. Distinctive:

    Do not use it to install a result, edit a Skill, or answer a general RDK question when the user has not asked for discovery.

    line 31
  4. Distinctive:

    Run the bundled search script before recommending any Skill or command:

    line 36
  5. Very specific:

    Read the JSON result. Each match is already ranked deterministically and includes its canonical name, description, catalog location, score, …

    line 42
  6. Very specific:

    If matches is empty, say that no registered RDK Skill matched and use the rdk-docs-reference fallback.

    line 49
  7. Very specific:

    If the script returns error, surface the index error and stop.

    line 52
  8. Very specific:

    Never execute an action automatically.

    line 58
  9. Distinctive:

    Treat an unreadable, malformed, or unsupported index as terminal.

    line 61

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

RDK Skill Finder

Purpose

Search the bundled, generated RDK catalog and return only registered Skills. It provides a Hub install command for flat Skills and hands workspace-integrated records to rdk-pack-installer.

When to use

Use when the user needs to discover the right RDK Skill or workspace pack from a task description, board/platform term, or exact name. This is especially useful for X5 toolchain, model quantization, compilation, diagnostics, and hardware workflow discovery.

Do not use it to install a result, edit a Skill, or answer a general RDK question when the user has not asked for discovery.

Instructions

  1. Run the bundled search script before recommending any Skill or command:

    python scripts/search_catalog.py QUERY [--pack NAME] [--platform TOKEN] [--install-type flat|workspace] [--limit N]
  2. Read the JSON result. Each match is already ranked deterministically and includes its canonical name, description, catalog location, score, and action. Use the first match unless the user provides a filter that asks for a different result.
  3. For a flat result, present the returned npx skills add command unchanged. For a workspace result, hand off to rdk-pack-installer; that Skill owns project-root confirmation and all installation writes.
  4. If matches is empty, say that no registered RDK Skill matched and use the rdk-docs-reference fallback. Do not invent a Skill, repository, command, or pack name.
  5. If the script returns error, surface the index error and stop. Do not guess from stale catalog data.

Safety

  • The finder is read-only: it only reads references/skill-index.json.
  • Never execute an action automatically. Installation requires the user's explicit request, and workspace installation must be handled by rdk-pack-installer with its confirmation gates.
  • Treat an unreadable, malformed, or unsupported index as terminal; do not substitute manually remembered catalog records.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free