From zie-framework
Load shared context bundle (ADRs + project context) once per session. Returns context_bundle for downstream reviewers.
How this skill is triggered — by the user, by Claude, or both
Slash command
/zie-framework:load-contexthaikuThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
<!-- FAST PATH -->
Purpose: Load ADR + project context bundle once per session for downstream reviewers. When to use fast path: context_bundle already provided → return it immediately. Quick steps: (1) If context_bundle provided → return it. (2) Else: cache check → disk fallback → return bundle.
Load ADRs and project context once. Returns context_bundle for every
downstream reviewer call in the session.
Fast-path: If context_bundle is provided as an argument to this skill
invocation → return context_bundle immediately. Skip all steps below.
Step 0: Load unified cache
get_cache_manager from hooks/utils_cache.py.cache = get_cache_manager(cwd) where cwd is project root.adrs_content = cache.get_or_compute("adrs", session_id, compute_fn, ttl=3600) where:
def compute_fn():
decisions_dir = cwd / "zie-framework" / "decisions"
adr_files = sorted(decisions_dir.glob("*.md"))
contents = [f.read_text() for f in adr_files if f.exists()]
return "\n\n".join(contents) if contents else ""
get_or_compute() returns cached value → skip disk read.get_or_compute() calls compute_fn() → reads ADRs from disk → caches result.Step 1: Design context
zie-framework/project/context.md →
context_content (empty string if file missing).
cache.get_or_compute("project_md", session_id, compute_fn, ttl=300) for caching.Step 2: Assemble bundle
context_bundle = {
adrs: adrs_content,
context: context_content
}
npx claudepluginhub zierocode/zie-plugins-claude-code --plugin zie-frameworkProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.