Skip to content

Repository files navigation

                                   _
  ___  ____ ___  ____  ____  _____(_)_  ______ ___
 / _ \/ __ `__ \/ __ \/ __ \/ ___/ / / / / __ `__ \
/  __/ / / / / / /_/ / /_/ / /  / / /_/ / / / / / /
\___/_/ /_/ /_/ .___/\____/_/  /_/\__,_/_/ /_/ /_/
             /_/

Claude Code plugin marketplace.

License: MIT

All heurema open-source AI agent tools. Craft, not conjuring.


Install

claude plugin marketplace add heurema/emporium

Then install any plugin:

claude plugin install signum@emporium
claude plugin install herald@emporium
claude plugin install arbiter@emporium
claude plugin install anvil@emporium
claude plugin install reporter@emporium
claude plugin install genesis@emporium
claude plugin install sentinel@emporium
claude plugin install teams-field-guide@emporium

Catalog

Plugin Description Repo
signum Risk-adaptive development pipeline with adversarial consensus code review signum
herald Daily curated news digest — zero API keys, fully local herald
arbiter Multi-AI orchestrator — Codex CLI + Gemini CLI for review, ask, implement, panel arbiter
anvil Plugin dev/test toolkit — scaffold, validate, test, review anvil
reporter Report bugs, request features, and ask questions for any heurema product reporter
sentinel AI workstation security audit — secrets, MCP, plugins, hooks, trust sentinel
genesis Evolve startup ideas through AI-powered natural selection with 3 agent personas genesis
teams-field-guide Comprehensive guide to Claude Code multi-agent systems teams-field-guide

Quality Infrastructure

Emporium includes tooling to enforce plugin quality before acceptance into the marketplace.

Architecture

lib/rubric/     12 pure check functions (no IO in core logic)
lib/trace/      Signum artifact archival with PII/secret redaction
scripts/        CLI tools for running checks
tests/          pytest test suite

CLI Tools

All scripts follow a common pattern: --plugin-root DIR (default: ~/personal/skill7/devtools/), --output FILE (JSON report), exit 0 = PASS, exit 1 = FAIL.

Ingestion Gate — 6 safety checks

python3 scripts/ingestion-gate.py --plugin herald
python3 scripts/ingestion-gate.py --plugin signum --output /tmp/report.json
python3 scripts/ingestion-gate.py --plugin arbiter --plugin-root /path/to/plugins

Checks: hardcoded_secrets, dangerous_commands, suspicious_urls, shell_true, line_count (300 max), required_sections.

Quality Rubric — 12-point scoring

python3 scripts/quality-rubric.py --plugin=herald
python3 scripts/quality-rubric.py --plugin=arbiter --output=/tmp/arbiter.json

4 structural + 4 safety + 4 quality checks. Score 0-12, gate PASS at >= 9.

Line Cap Enforcer — batch 300-line scanner

python3 scripts/enforce-line-cap.py
python3 scripts/enforce-line-cap.py --plugin-root /path/to/plugins --output /tmp/linecap.json

Scans all plugins, reports files exceeding 300 lines.

Failure Modes Checker — required SKILL.md sections

python3 scripts/check-failure-modes.py
python3 scripts/check-failure-modes.py --plugin-root /path/to/plugins

Verifies every plugin SKILL.md has description, usage, and known failure modes sections.

Trace Archive — Signum artifact archival

python3 scripts/trace-archive.py --archive
python3 scripts/trace-archive.py --prune --max-age 30
python3 scripts/trace-archive.py --archive --prune

Archives .signum/ run artifacts with automatic redaction of secrets, API keys, emails, and IP addresses. Prunes old archives by age.

Library API

lib/rubric — Check Functions

from lib.rubric import (
    # Structural
    check_skill_md,           # SKILL.md exists and non-empty
    check_required_sections,  # description + usage + failure modes
    check_line_count,         # no file > 300 lines
    check_skill_id_format,    # valid skill_id pattern

    # Safety
    check_hardcoded_secrets,  # no embedded credentials
    check_no_dangerous_commands,  # no rm -rf, eval, etc.
    check_no_suspicious_urls,     # no non-HTTPS URLs
    check_no_shell_true,          # no shell=True in subprocess

    # Quality
    check_has_bench_tasks,    # bench/ tasks exist
    check_has_examples,       # usage examples present
    check_semver,             # valid semver version
    check_metadata_complete,  # author, description, category
)

# Every function: (Path) -> (bool, str)
passed, detail = check_hardcoded_secrets(Path("/path/to/plugin"))

lib/trace — Archive & Redaction

from lib.trace import redact_text, archive_signum_run, prune_old_runs

# Redact secrets/PII from text
clean = redact_text("key=sk-1234567890abcdef, email: user@example.com")

# Archive .signum/ with redaction (returns Path to archived run)
archive_path = archive_signum_run(".signum", "~/.local/share/emporium/traces")

# Delete runs older than 90 days
deleted = prune_old_runs("~/.local/share/emporium/traces", max_age_days=90)

Tests

cd ~/personal/heurema/emporium
python3 -m pytest tests/ -v

8 tests covering archive, redaction, pruning, file integrity, and error handling.

Adding a New Plugin

See CONTRIBUTING.md.

See Also

  • signum — risk-adaptive development pipeline with adversarial code review
  • herald — daily curated news digest plugin for Claude Code
  • arbiter — multi-AI orchestrator (Codex + Gemini)
  • anvil — plugin dev/test toolkit
  • reporter — issue reporter for heurema products
  • genesis — evolve startup ideas with memetic algorithm and 3 AI agent personas
  • sentinel — AI workstation security audit
  • teams-field-guide — comprehensive guide to Claude Code multi-agent teams
  • proofpack — proof-carrying CI gate for AI agent changes

License

MIT

About

Plugin marketplace for heurema tools. Craft, not conjuring.

Topics

Resources

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages