hforge-recursive-investigate

Automate recursive investigation workflows for ambiguous cross-module tasks.

3|Updated Mar 24, 2026
One-click install
npx skills add https://github.com/ldilov/harness-forge --skill hforge-recursive-investigate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hforge-recursive-investigate
Source: https://github.com/ldilov/harness-forge/tree/main/.agents/skills/hforge-recursive-investigate
Command: npx skills add https://github.com/ldilov/harness-forge --skill hforge-recursive-investigate

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Auto-discoverable wrapper that coordinates Harness Forge's recursive-investigate workflow for ambiguous, cross-module tasks, delivering structured plans and durable recursive artifacts.

Core Features & Use Cases

  • Auto-discovery and orchestration of recursive investigations across modules.
  • Support for Typed RLM with bounded subcalls and durable artifact tracking.
  • Session planning, plan execution guidance, and artifact replay for reproducibility.
  • Works best in repositories initialized with Harness Forge.

Quick Start

Run the recursive-investigate workflow on the active task to generate a structured plan and artifacts.

Frequently Asked Questions about hforge-recursive-investigate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I plan cross-module software investigations that require staged reasoning?

Recursive investigation workflows automate cross-module software planning by applying staged reasoning to ambiguous tasks, generating structured plans and durable artifacts for multi-module execution. This coordinates bounded subcalls across modules to produce reproducible session plans.

What is a recursive investigation workflow for ambiguous software engineering tasks?

A recursive investigation workflow is an automated process that breaks down ambiguous, cross-module software engineering tasks into staged reasoning steps, tracking artifacts durably across sessions to produce structured execution plans and reproducible results.

Does recursive investigation work with repositories not initialized with Harness Forge?

Recursive investigation works best in repositories initialized with Harness Forge, as it relies on auto-discovery and orchestration features specific to that environment to coordinate bounded subcalls and durable artifact tracking effectively.

How do I track durable artifacts across multiple investigation sessions?

Durable artifact tracking across sessions is handled automatically by the recursive investigation workflow, which generates and persists structured artifacts during staged reasoning, enabling session planning, execution guidance, and artifact replay for later reproducibility.

When should I use a recursive investigation workflow instead of manual planning?

Use recursive investigation workflows when tasks span multiple modules and require staged reasoning, durable artifact tracking, or cross-session reproducibility; manual planning is insufficient for ambiguous, multi-module software engineering tasks needing bounded subcalls and structured plans.