
from stellarlinkco
A framework for long-running, multi-session agent tasks: progress checkpointing, failure recovery, task dependency management, and concurrency controls.
A comprehensive protocol for running agent workloads that span multiple sessions. The harness provides file-based progress persistence, atomic locking, checkpointing, dependency-aware task selection, and recovery rules for interrupted work. It defines lifecycle commands (/harness init, run, status, add) and explicit validation/rollback behaviors.
Use this skill when a job requires progress to survive context resets, distributed workers, or when tasks have dependencies and need reliable retries and rollbacks (e.g., long CI jobs, multi-step migrations, batch automation across sessions). Triggers on harness-related commands or when multi-session orchestration is requested.
Best used by agents that can read/write files, run shell/git commands, and manage isolated working directories (CLI-capable agents like Copilot/Codex/Gemini CLI).
A thorough long-running agent task framework implementing progress checkpointing, failure recovery, dependency management, and concurrency controls. The SKILL.md is exceptionally well-structured with detailed protocols, decision matrices, and error handling strategies. No bundled scripts — the skill is a pure protocol specification requiring only bash and git. The only notable concern is `git reset --hard` and `git clean -fd` usage, which are intentional rollback mechanisms properly scoped to task commits.
Impressively comprehensive skill document. The dual-file persistence model (append-only log + structured JSON) with atomic writes and backup protocol is well-designed. Lock mechanism using mkdir is portable and handles stale locks. No security issues found — all potentially destructive commands are properly scoped and intentional.
Do — Feature Development Orchestrator
Orchestrates a five-phase, multi-agent feature development workflow (Understand, Clarify, Design, Implement, Complete) using codeagent-wrapper; manages task wor
Codeagent Wrapper
Run and orchestrate codeagent-wrapper tasks across multiple AI backends (Codex, Claude, Gemini, OpenCode) for parallel code analysis, large refactors, and multi
OmO - Multi-Agent Orchestrator
Routing-first orchestrator that invokes and coordinates specialized agents for code analysis, bug investigation, planning, and implementation.