Skip to main content
Glama

draft_skill

Draft a deterministic SKILL.md for a suggested skill candidate with triggers, example commands, failure modes, verification, and evidence. Review it before installing.

Instructions

Draft an Agent Skills SKILL.md (works in Claude Code, Codex, Gemini CLI, Cursor, OpenCode...) for a candidate from suggest_skills.

The draft is deterministic: frontmatter (name, description with triggers), steps with real example commands, known failure modes and their fixes, verification, and evidence. Review and edit it, then pass it to install_skill. Use get_session on the evidence sessions to refine it.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "candidate_id": "<candidate_id>"
    },
    "name": "draft_skill"
  }
}

Input Schema

NameRequiredDescriptionDefault
sourcesNoall
candidate_idYes

Output Schema

NameRequiredDescriptionDefault
resultYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.4.0

TDQS

A4.3/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the burden, and it does disclose that the draft is deterministic and enumerates what it contains (frontmatter with triggers, example commands, failure modes and fixes, verification, evidence). What is left unstated is whether anything is persisted to disk or whether the output is purely returned content, and whether a candidate_id must exist before calling.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three tight sentences: purpose and artifact first, then a content breakdown, then the next-step routing. No filler, and the most load-bearing information is front-loaded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists so return values need no explanation, and the surrounding workflow is fully described. The only meaningful omission is the unexplained 'sources' parameter, which matters for a tool whose whole output depends on which sessions it reads.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters2/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 0%, so the description must compensate, and it only partially does: 'for a candidate from suggest_skills' clarifies that candidate_id references an upstream suggestion. The 'sources' parameter (array-or-string, default 'all') is never mentioned, leaving one of two parameters entirely unexplained.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and artifact: draft a SKILL.md file for Agent Skills, targeting a candidate produced by suggest_skills. It also names the runtime targets (Claude Code, Codex, Gemini CLI...), so the agent knows exactly what is being produced and for where.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Describes the full pipeline: consume a candidate from suggest_skills, draft, review and edit, then pass to install_skill. It also adds a refinement path via get_session on the evidence sessions, giving explicit alternatives for the follow-up steps rather than leaving them to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.