Delivery Harness Framework / DHF Beginner Guide

What DHF does before an agent starts building.

DHF is a delivery framework for agentic engineering. It turns a broad request into recovered facts, lifecycle routing, scoped execution, verification evidence, and a handoff that the next session can resume.

Plain Definition

What DHF Does

DHF answers five questions before it claims progress: what facts were read, which lifecycle stage the work is in, which skill or helper owns the next step, how the result will be verified, and what the next safe task is.

Recover

Facts first

Read repo instructions, durable state, git status, and recent evidence before trusting chat memory.

Route

Stage-aware work

Choose requirements, planning, development, validation, review, ship, or handoff before choosing tools.

Prove

Evidence over vibes

Completion needs command, exit_code, key_output, and timestamp.

Five-step flow

Five-step flow

1
Recover

Read durable repo facts and restore phase, dirty state, latest verification, and next safe task.

2
Clarify

If the goal, scope, vocabulary, or acceptance criteria are unclear, stop and lock requirements first.

3
Route

Select the lifecycle owner: planner, TDD, QA, security review, ship workflow, docs, or a runtime helper.

4
Verify

Run the relevant command and capture the evidence fields before saying the work is complete.

5
Handoff

Record changed surfaces, verification, blockers, and the next safe task for the next agent.

Concrete IO

Input request -> DHF output

input request

"Improve the public DHF docs so an English reader knows where to start, and prove the links are valid."

DHF output

  • stage: validation
  • command: python3 test_runner.py
  • exit_code: 0
  • key_output: [PASS] lifecycle skill routing doc discoverable
  • timestamp: 2026-05-14T00:00:00-04:00
Next

Continue the English path