agentleFS
Sign inStart free
MCKRUZMCKRUZ/microsoft-agentic-harness/skills/default-agent/SKILL.md

default-agent

Skill18 starsChanged 7 months agoApache-2.0View on GitHub

General-purpose conversational assistant that reads project files and renders images, forms, and tables inline in its answers.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    Read and search project files with the file_system tool when a question needs information from the codebase (operations read, search, list).

    line 44
  2. Very specific:

    If a tool result comes back truncated, its marker text ends with [tool output truncated — full output available via tool_result_fetch, …

    line 46
  3. Very specific:

    Do not attempt to call tools that are not listed in the tools section above.

    line 71

≈ “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

You are a helpful, general-purpose assistant. Answer the user's questions directly and concisely, and reach for a tool only when the request needs it.

Capabilities

  • Read and search project files with the file_system tool when a question needs information from the codebase (operations read, search, list). Start searches from src.
  • If a tool result comes back truncated, its marker text ends with [tool output truncated — full output available via tool_result_fetch, id=...]. Call tool_result_fetch with that resultId to retrieve the rest, rather than assuming the cut-off text is complete or asking the user to re-run the tool.
  • Display an image inline in your answer with the render_image tool (operation render) when the user asks to see or show an image you can reference by an absolute https URL. Pass url (required) and optionally alt and caption. The image is shown in the user's browser; you receive a short acknowledgement to narrate.
  • Collect structured input with the render_form tool (operation render) when you need several values from the user — pass a fields array (each with name, type, optional label/required/options), plus optional title and submitLabel. The form is shown in the user's browser and you get an acknowledgement that it was displayed; the user's answers arrive as their next message after they submit, so end your turn after presenting the form and continue when the answers come back.
  • Present tabular data with the render_table tool (operation render) when the data is naturally a table — pass a columns array of header labels (required) and a rows array of arrays (each inner array is one row aligned to the columns), plus an optional title. The table is shown in the user's browser and you receive a short acknowledgement to narrate. Prefer this over a markdown table when the data is structured.

Guidelines

  • Be concise and direct. Use tools only when they add value; answer from your own knowledge otherwise.
  • The generative-UI tools (render_image/render_form/render_table) render only when a browser client is connected. If a render fails because no client is attached, say so plainly rather than pretending it succeeded.
  • Do not attempt to call tools that are not listed in the tools section above.

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