By caasi
Define and validate multi-step agent workflows using an Arrow-style DSL with operators for sequencing, branching, looping, and variable binding. Allows syntax validation and tool call planning before execution.
A Claude Code plugin marketplace by caasi.
Delegate tasks to external OpenAI-compatible chat endpoints with capability probing, delegation patterns, and prompt injection awareness.
See plugins/chat-subagent/skills/chat-subagent/README.md for full documentation.
Describe multi-step agent workflows using an Arrow-style DSL and validate them structurally with the ocaml-compose-dsl binary. The DSL is a planning language — the agent expands it into concrete tool calls.
See plugins/compose/skills/compose/README.md for full documentation.
Reflect on human-AI stewardship through Socratic dialogue. A mirror, not a checklist — rooted in Audrey Tang's Humane Intelligence framework and Civic AI 6-Pack of Care.
See plugins/kami/skills/kami/README.md for full documentation.
claude plugin marketplace add caasi/dong3
MIT
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub caasi/dong3 --plugin composeAssisted review loop — local reviewers answer blind and in parallel, each returns its own convergence verdict (loop-until-dry, facilitator owns K + escalation); roster enrolled per host via /review-loop:init; never merges autonomously; ships an always-on UserPromptSubmit hook that writes nothing until you opt in via /review-loop:init
Git-native preparation & convergence — four routines: init/prepare/converge/prune. prepare is local-first: work stays on local prepare/* by default; remote push is a cross-machine opt-in (push-prepare-branches: yes); prepare freshness-rebases in-progress branches onto default (schema 7; opt-out rebase-prepare-onto-default), and converge offers the refresh. Human takeover recognized by containment; redundant prepare/<slug> surfaced for cleanup, never auto-deleted. converge accept defaults to a handoff-only rename (with a human-marked maintenance exception); prune sweeps unused local + remote branches (human-present, read-only until approved). Committed .tsugu/ is a WIP-knowledge layer (policy.md + context.md + knowledge/) plus a one-line .gitattributes (context.md merge=union); state derived from refs (single-layer, no intake notes); never auto-merges; invokes no user-installed skill by default.
Platform-specific fetch strategies for content that resists simple WebFetch
Delegate tasks to external chat endpoints (OpenAI-compatible and LM Studio native API) as a subagent
Socratic dialogue for reflecting on human-AI stewardship
Harness-native ECC plugin for engineering teams - 67 agents, 279 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).
Claude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Plugin-safe Claude Code distribution of Agentic Awesome Skills with 1,933 supported skills.