lilpacy/claude/skills/claude-skill-reclassify-agents
claude/skills/claude-skill-reclassify-agents/SKILL.md
# Skill: Reclassify existing subagents vs skills and migrate ## Intent Scan an existing repo for *Claude Code subagents* and *skills*, decide which each item should be, and **apply the conversion** by moving/rewrapping files into the appropriate structure — with backups and a migration report. This skill is designed for repos that contain: - subagents as Markdown files (commonly `.claude/agents/*.md`) - skills as either: - folder-based skills (`.claude/skills/<skill>/SKILL.md`), or - singl
$ install --global
skillsauthnpx skillsauth add lilpacy/dotfiles claude/skills/claude-skill-reclassify-agentsInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 8, 2026, 6:14 AM158.6s10 files scanned
SKILL.md
Skill: Reclassify existing subagents vs skills and migrate
Intent
Scan an existing repo for Claude Code subagents and skills, decide which each item should be, and apply the conversion by moving/rewrapping files into the appropriate structure — with backups and a migration report.
This skill is designed for repos that contain:
- subagents as Markdown files (commonly
.claude/agents/*.md) - skills as either:
- folder-based skills (
.claude/skills/<skill>/SKILL.md), or - single Markdown skills (various conventions)
- folder-based skills (
If your structure differs, adjust config.yaml.
Inputs you need from the user (or derive automatically)
- repository root (assume current working directory is repo root)
- optional: path to config (
config.yaml) - desired mode:
dry-run(generate report only)apply(convert + update references)
Outputs
reclassify-report.md- list of items
- current type (agent/skill/unknown)
- recommended type
- confidence score
- top rationale signals
- migration plan summary
- backups under
.reclassify_backup/<timestamp>/... - migrated files in:
target_agents_dir(default.claude/agents)target_skills_dir(default.claude/skills/<name>/SKILL.md)
Operating rules (guardrails)
- Always create a backup snapshot under
.reclassify_backup/...before writing. - Prefer running on a git branch (create one if possible).
- Never delete originals; if moving, keep an embedded appendix OR keep an original copy under backup.
- If an item references tools/paths that do not exist, preserve the text; do not “fix” behavior beyond migration.
- Update references best-effort (rename mentions), but do not break code; keep replacements conservative.
Procedure (high-level)
- Discover existing agents & skills by scanning configured directories (and auto-detect when missing).
- Classify each item using the rubric:
- Skill score ↑ with procedural/template/checklist/format markers
- Subagent score ↑ with role/principles/judgment/exploration markers
- Report: write
reclassify-report.mdwith confidence & rationale. - Convert:
- Agent → Skill:
- create folder
target_skills_dir/<name>/SKILL.md - rewrap content into a Skill template (Goal / Inputs / Outputs / Procedure / Guardrails)
- create folder
- Skill → Agent:
- create
target_agents_dir/<name>.md - rewrap content into an Agent template (Role / Principles / How to respond)
- create
- Agent → Skill:
- Reference updates (optional):
- search Markdown files and replace references to moved paths/names where safe.
- Summary: print a small summary and point to report + backup path.
How to run (recommended)
Use the provided script to execute the procedure end-to-end:
python3 .claude/skills/reclassify-agents/scripts/reclassify.py --config .claude/skills/reclassify-agents/config.yaml --mode dry-run
python3 .claude/skills/reclassify-agents/scripts/reclassify.py --config .claude/skills/reclassify-agents/config.yaml --mode apply
Related Skills
lilpacy/web-doc-reading
development
VerifiedTrustedCommunity
Use when searching the web or reading online documentation. Prefer DuckDuckGo for search and read documents through npx curl.md instead of raw HTML.
1SKILL.mdUpdated May 30, 2026
lilpacy/japanese-test-conventions
testing
VerifiedTrustedCommunity
Use when writing or editing tests. Tests should be ordered by near-normal, normal, then abnormal cases where applicable, and test names must be Japanese behavior descriptions from a reviewer/user perspective.
1SKILL.mdUpdated May 30, 2026
lilpacy/gof-functional-patterns
development
VerifiedTrustedCommunity
GoF/オブジェクト指向デザインパターンを関数型プログラミング（pure functions, higher-order functions, ADT, composition, immutability, effect boundaries）でシンプルに整理・設計・リファクタリングする。Strategy/Factory/Adapter/ObserverなどGoF全23パターンのFP置き換え、適用判断、具体事例を提示する必要があるときに使う。
1SKILL.mdUpdated May 30, 2026
lilpacy/git-commit-workflow
tools
VerifiedTrustedCommunity
Use when committing, pushing, or preparing PRs. Defines the user's commit workflow, message style discovery, review handoff, and branch/worktree push requirements.
1SKILL.mdUpdated May 30, 2026