MCP App StoreMCP App Store
from orchestrate17
Launch and manage agent runs: compose prompts, route models to the correct CLI, log artifacts, and inspect run history.
Run-Agent provides a single CLI entrypoint to compose and execute agent runs. A run is model + optional agent profile + optional skills + prompt. The skill routes chosen models to the appropriate runtime (Claude, Codex, OpenCode), launches runs, records structured artifacts (params, prompts, outputs, logs) and maintains an append-only run index for inspection and troubleshooting.
Use Run-Agent when you need reproducible agent executions, to orchestrate subagent runs, or to route tasks across different model CLIs. It is ideal for automating repeatable workflows, running dry-runs to preview prompts, grouping runs by session, and debugging failed runs.
.orchestrate/ layout (has_references=true).Works with standard CLI-harness agents that use model CLIs such as Claude, Codex, and OpenCode (agent-level orchestrators that call out to these CLIs).
Run-agent is a multi-harness agent execution engine that routes models to claude/codex/opencode CLIs, composes prompts from skills, and logs structured run artifacts. Scripts are well-written with proper error handling and set -euo pipefail, but the main run-agent.sh depends on unbundled lib/ modules (parse.sh, prompt.sh, logging.sh, exec.sh) that weren't included. Utility scripts (extract-*, log-inspect.sh) all failed gracefully with usage messages when run without args, which is expected behavior. No security concerns beyond minor shell variable quoting risks in run-index.sh label filtering.
jqlib/parse.sh (not bundled)Well-engineered orchestration skill with clean architecture. Main concern is completeness: the core run-agent.sh depends on 4 lib/ modules not bundled in the skill. The label filtering in run-index.sh uses variable interpolation in jq select expressions which could be a minor injection risk if labels contain special characters, but in practice this is low severity. All scripts handle errors and missing args gracefully.
Development Worktree
Create an isolated git worktree for feature work, auto-run project setup, and verify a clean test baseline before development.
ds-fix — data-science mid-analysis fixer
Orchestrates diagnosis and targeted fixes mid-analysis: diagnose root cause, apply fixes with output-first verification, and update project learnings.
Readwise Reader Document Management
Manage Readwise Reader documents: list, save, search, move, tag, highlight, export and bulk-edit via official and custom CLIs.
Bounty Hunter — Atlas
Persona skill: 'Atlas' — a profit-focused developer persona for discovering, evaluating and executing paid bounties or freelance tasks with ROI-aware workflows.
Junshi — Research Advisor
Daily strategic research advisor that scans arXiv/venues, digests papers, and proposes bold, ranked research ideas tailored to the user's profile.
Full Stack Builder
End-to-end builder that scaffolds, implements, tests, and optionally deploys web and API applications from a natural-language specification.
ezBookkeeping API Tools
Command-line API tools for ezBookkeeping: record and query transactions, retrieve accounts/categories/tags, and fetch exchange rates for self-hosted personal fi
Feishu Voice Sender
Convert MP3s and send them as native Feishu voice messages (playable voice clips) to users or groups.