Claude-skill-registry create-agent-sop

Create agent-ready Standard Operating Procedure (SOP) skills that autonomous agents can execute deterministically. Use when creating workflow skills, procedural skills, or any skill that needs to be unambiguous, idempotent, robust, and observable. Triggers on "create SOP", "new workflow skill", "agent-ready procedure".

install

source · Clone the upstream repo

git clone https://github.com/majiayu000/claude-skill-registry

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/create-agent-sop" ~/.claude/skills/majiayu000-claude-skill-registry-create-agent-sop && rm -rf "$T"

manifest: skills/data/create-agent-sop/SKILL.md

Create Agent SOP

Meta-skill for generating SOP skills that autonomous agents can execute without human intervention.

When to Use

Creating workflow skills (e.g., feature-development-loop, weekly-review)
Procedural skills with clear start/end states
Skills requiring deterministic, repeatable execution

Process

1. Gather Requirements

Ask user for:

Workflow name (hyphen-case, e.g.,
```
feature-development-loop
```
)
Trigger phrases (when should this SOP activate?)
Steps (what are the major phases?)
Success criteria (how do we know it's done?)

2. Initialize Skill

pnpm exec tsx .claude/skills/create-agent-sop/scripts/init_sop_skill.ts <name> --path .claude/skills

3. Write SKILL.md

Follow the template in

references/sop-template.md

. Key sections:

---
name: <skill-name>
description: <what + when to use>
---

# <Skill Name>

## Prerequisites
[Tools, permissions, state requirements]

## Input Schema
[Typed parameters with defaults]

## Procedure
[Numbered steps with @command tags]

## Success Criteria
[Verification checks]

## Error Recovery
[Fallback paths]

4. Add Tools (Optional)

Before adding scripts, evaluate if the functionality should be a CLI tool instead:

Consider CLI Tool When	Use Script When
Multiple clients could use it	Skill-specific logic only
Reusable across skills	Simple validation/checks
Complex logic worth maintaining	Glue code between tools
Would simplify the script significantly	One-off operations

CLI tools go in

packages/cli/

and are invoked via

sp

or tool-specific commands. Skills can then call these tools, keeping scripts minimal.

Scripts go in

scripts/

for skill-specific operations:

scripts/validate.sh      # Prerequisite checks
scripts/check-state.ts   # Skill-specific state validation

5. Validate

Run the validation script to check against the 10 SOP characteristics:

pnpm exec tsx .claude/skills/create-agent-sop/scripts/validate_sop_skill.ts <skill-path>

Output shows pass/fail for each characteristic. See

references/sop-characteristics.md

for details.

6. Package

Create a distributable

.skill

file:

pnpm exec tsx .claude/skills/create-agent-sop/scripts/package_sop_skill.ts <skill-path> [output-dir]

Validates first, then creates

<skill-name>.skill

(zip format).

References

```
references/sop-characteristics.md
```
- The 10 characteristics explained
```
references/sop-template.md
```
- SKILL.md template for SOPs