skillgen

A guided process for turning a repeated workflow into a reusable agent skill. It analyzes the pattern, creates the skill files, checks them for security, and activates them only with human approval.

In plain words
What is it for?
Use it to create skills from a description or detected recurring workflow, resume the process at a particular stage, or move to a named stage.
Why use it?
It prevents skills from being created or enabled automatically without review. The approval stages make the generated instructions easier to check and control.

Cursor rule for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/redhat-vmeperf/agent-tasks-template/skillgen
Clone the repo
git clone --depth 1 https://github.com/redhat-vmeperf/agent-tasks-template

Made for: Cursor.

Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 374 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00031 $0.00374
Opus 5 $0.00015 $0.00187
Sonnet 5 $0.00006 $0.00075
Haiku 4.5 $0.00003 $0.00037

Measured 2d ago against content hash b2e157fadc9f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

skillgen scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/rules/skillgen.mdc · 33 lines

What it actually says

Skill Generation Pipeline

This rule runs the 4-gate Skill Generation Pipeline for creating new skills from detected workflow patterns. Read .agents/pipelines/SKILLGEN.md for the full pipeline process definition, gate definitions, escalation protocol, and all process rules.

Invocation

  • "Generate a skill for [description]" — Start the pipeline from Gate 1 with a manual skill description
  • "Resume skillgen at gate N for [skill]" — Resume the pipeline at Gate N
  • "Jump to [gate name] for [skill]" — Jump to a specific gate (analyze, generate, security, activate)

If the user does not provide a skill description, ask them to describe the skill or the recurring workflow pattern they want to codify.

Skill Generation Context

This pipeline produces skill definitions, pipeline definitions, role files, and platform entry points as its output, not code.

  • Gate 1 blocks: No skill is generated until a human confirms the detected pattern
  • Gate 2 output: Complete skill file set (SKILL.md, PIPELINE.md, roles, platform entries) + Generation Report
  • Gate 3 blocks: No skill activates without passing security review
  • Gate 4 final action: WRITE FILES + ACTIVATE SKILL (human-approved)

Security

  • Security steps cannot be skipped.
  • No interaction with any service or system may reduce the security posture of data passing through the process (REQ-5).
  • Generated skill files must pass the full SECURITY_REVIEW_CHECKLIST.md before activation (REQ-6 equivalent).
  • Generated skills are treated as untrusted content until human-approved.
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 33 lines · 31 tokens per session scan A b2e157fadc9f

Subscribe to this mod's changes

skillgen is a cursor rule published in the GitHub repository redhat-vmeperf/agent-tasks-template (3 stars, last pushed 4mo ago), licensed Unlicense. It adds 31 tokens to every session and 374 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.