▲ a toolkit for the Claude Code CLI

THE SKILLHARNESS

Prompt-powered /skills that carry a change from brainstorm to ship. Each skill does one step and hands off to the next.

see the pipeline
13
skills
the engine

What a skill is

A skill is a prompt program that does one step of a change. It has a defined scope and output, and hands off to the next skill. You invoke each one as /skill-name in Claude Code.

the full chain

brainstorm → ship

Each skill hands off to the next. /feature-intake runs first and routes the change to the steps it needs.

/feature-intakeruns first
Classify the request — 10-flag risk checklist + hard gates → a lane, plus a confidence score.
specs/<slug>/SUMMARY.md (Lane / Confidence / route)
/brainstorminghigh-risk lane
Explore intent, requirements & design through dialogue — before any code.
specs/<slug>/design.md
/xia2research
Investigate what already exists — locally, upstream on GitHub, and in version-matched docs — before building.
specs/<slug>/research-brief.md
/writing-plansplan
Turn the design into XML tasks grouped into waves (parallel-safe, each with a <verify> command).
specs/<slug>/PLAN.md
/visual-plannerauto
A deterministic script renders the plan to a self-contained HTML for visual review, then opens it.
specs/<slug>/PLAN.html
/using-git-worktreesisolate
Create an isolated worktree + branch so feature work never touches the main workspace.
worktree + branch
/subagent-driven-developmentbuild
Execute the plan task-by-task — a fresh subagent per task, one read-only reviewer per task returning two independent verdicts (spec compliance + code quality), then delegates the whole-diff review passes below.
implemented + reviewed tasks
/context-propagation-auditconditional
Only when the diff touches workflow-as-code (skills, agents, rules). Enumerates every consumer and execution context and proves each one actually receives the instruction — assumed delivery fails.
PASS/FAIL per consumer
/correctness-reviewreview
Adversarial runtime-bug hunt over the whole diff — find → score (0–100 per finding, threshold 75) → classify → fix-loop. Invokable standalone on any diff.
fixes or escalations
/intent-reviewfinal pass
The last gate before shipping. Compares the diff against the original request — deliberately blind to PLAN.md, so a plan that drifted from what you asked for can't launder the drift.
intent verdict or escalation
/compoundlearn
Crystallize any non-obvious bug fix, pattern, or decision into discoverable knowledge.
docs/solutions/<category>/<slug>.md
/finishing-a-development-branchship
Verify the branch, pin the required review receipts, push, and open a PR — never merges. A human reviews and merges.
🚀 shipped
ls skills/

The skill library

Each skill is a Markdown prompt program with a defined scope, hard gates, and a handoff. Click one to expand.

get started

Drop it into your repo

Skills are loaded by the Claude Code CLI. Add them to an existing project with one line.

01 · Add the skills to a project

Copies the skills into .claude/. Needs git + jq. Existing files are backed up first.

# adds the skills & agents (the installer also copies hooks & rules)
curl -fsSL "https://raw.githubusercontent.com/minhtran3124/agent-harness/main/scripts/install-harness.sh?$(date +%s)" | bash -s -- --yes

02 · Start a change

Begin with intake. It classifies the change and routes it to the next skill.

/feature-intake   # classify → route to the next skill