# Complex Task Protocol

> Auto-discoverable wrapper for `.hforge/library/skills/complex-task-protocol/SKILL.md`.

- Skill: `ldilov/complex-task-protocol-2` (Agent Skill)
- Install (CLI): `npx skillmds@latest add ldilov/complex-task-protocol-2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ldilov/complex-task-protocol-2/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Productivity
- Author: ldilov (https://skillmd.com/u/ldilov)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/ldilov/complex-task-protocol-2

---


# Complex Task Protocol

## Activation

- trigger when a task is multi-step, risky, cross-module, orchestration-heavy, or likely to need verification and recovery discipline
- trigger when the task may benefit from a small number of bounded subagents for independent sidecar work
- trigger when reusable learning may need to be captured after complex work

## Use These Surfaces

- `.hforge/library/skills/complex-task-protocol/SKILL.md`
- `.hforge/runtime/agent-brief.md`
- `.hforge/generated/agent-command-catalog.json`
- `.hforge/runtime/recursive/language-capabilities.json`

## Operating Rule

Use the canonical skill under `.hforge/library/skills/` for execution. Treat this wrapper as a discovery entrypoint only. Activate the protocol by task complexity thresholds rather than by requiring a slash command.