claude-skills/scripts/check_plugin_json.py
Claude f47972967a
chore(phase-0): add dual-publish + plugin.json validation tools
Phase 0 of the multi-skill build: ship the two stdlib-only tools that
the rest of the work depends on.

- scripts/sync_skill_bundles.py: mirror a standalone plugin's
  SKILL.md + scripts/ + references/ + assets/ into its domain-bundled
  location. --check exits 1 on drift; --sync rewrites the mirror.

- scripts/check_plugin_json.py: validate plugin.json against the
  strict ClawHub schema (exactly the 8 allowed fields, semver version,
  author{name,url}, skills as string or array — bare "./" rejected per
  Claude Code v2.1.107+).

Verified: --all run reports OK on all 30 existing plugin.json files;
sync --check correctly detects missing mirrors. Karpathy-coder gate:
both files score 85/100 under strict (single nesting-depth WARN, no
FAIL) — better than the canonical karpathy-coder tools themselves.

https://claude.ai/code/session_01Dq12xJakFRxwaoU8Pqejdm
2026-05-09 05:54:15 +00:00

140 lines
4.2 KiB
Python
Executable file

#!/usr/bin/env python3
"""Validate plugin.json files against the strict ClawHub schema.
Required fields (exactly these 8, no others):
name, description, version, author{name,url}, homepage, repository, license, skills
skills: must be either a string ("./skills") or an array of relative paths.
The bare "./" form is REJECTED (Claude Code v2.1.107+ rejects it).
"""
import argparse
import json
import os
import re
import sys
REPO = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
ALLOWED = {"name", "description", "version", "author", "homepage", "repository", "license", "skills"}
STRING_FIELDS = ("name", "description", "homepage", "repository", "license")
SEMVER = re.compile(r"^\d+\.\d+\.\d+(?:-[\w.]+)?$")
def _check_keys(data):
keys = set(data.keys())
errors = []
extra = keys - ALLOWED
missing = ALLOWED - keys
if extra:
errors.append(f"extra fields: {sorted(extra)}")
if missing:
errors.append(f"missing fields: {sorted(missing)}")
return errors
def _check_strings(data):
return [f"{k}: must be string" for k in STRING_FIELDS if k in data and not isinstance(data[k], str)]
def _check_version(data):
if "version" not in data:
return []
v = data["version"]
if not isinstance(v, str) or not SEMVER.match(v):
return [f"version: must match semver, got {v!r}"]
return []
def _check_author(data):
if "author" not in data:
return []
a = data["author"]
if not isinstance(a, dict):
return ["author: must be object {name, url}"]
errors = []
if not isinstance(a.get("name"), str):
errors.append("author.name: must be string")
if not isinstance(a.get("url"), str):
errors.append("author.url: must be string")
extra = set(a.keys()) - {"name", "url"}
if extra:
errors.append(f"author: extra fields {sorted(extra)}")
return errors
def _check_skills_string(s):
if s in ("./", ""):
return ['skills: "./" is rejected by Claude Code v2.1.107+; use "./skills" or an array']
return []
def _check_skills_array(s):
if not s:
return ["skills: array is empty"]
errors = []
for entry in s:
if not isinstance(entry, str):
errors.append(f"skills: entries must be strings, got {entry!r}")
elif entry == "./":
errors.append('skills: "./" is rejected by Claude Code v2.1.107+; list explicit subfolders')
return errors
def _check_skills(data):
if "skills" not in data:
return []
s = data["skills"]
if isinstance(s, str):
return _check_skills_string(s)
if isinstance(s, list):
return _check_skills_array(s)
return ["skills: must be string or array of strings"]
def validate(path):
try:
with open(path) as f:
data = json.load(f)
except (OSError, json.JSONDecodeError) as e:
return [f"unreadable JSON: {e}"]
return (_check_keys(data) + _check_strings(data) + _check_version(data)
+ _check_author(data) + _check_skills(data))
def find_all():
out = []
for root, dirs, files in os.walk(REPO):
if any(skip in root for skip in (".git", "node_modules", "eval-workspace", ".gemini")):
dirs[:] = []
continue
if "plugin.json" in files and root.endswith(".claude-plugin"):
out.append(os.path.join(root, "plugin.json"))
return sorted(out)
def main():
ap = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
g = ap.add_mutually_exclusive_group(required=True)
g.add_argument("path", nargs="?", help="Path to a plugin.json file")
g.add_argument("--all", action="store_true", help="Validate every plugin.json in the repo")
args = ap.parse_args()
targets = find_all() if args.all else [args.path]
failed = 0
for t in targets:
errs = validate(t)
rel = os.path.relpath(t, REPO)
if errs:
failed += 1
print(f"FAIL {rel}")
for e in errs:
print(f" - {e}")
else:
print(f"OK {rel}")
if failed:
print(f"\n{failed} file(s) failed validation", file=sys.stderr)
return 1
return 0
if __name__ == "__main__":
sys.exit(main())