From harness-kit
Research / feasibility mode. Light-weight requirement analysis + a free-form findings document. No design, no Gate Review, no code. Use when "can we even do X?" is the question and you need evidence, not a vetted plan. Skips most of the 7-agent pipeline by design.
How this skill is triggered — by the user, by Claude, or both
Slash command
/harness-kit:harness-exploreThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Lightweight feasibility / research mode. Spend ~10-30 minutes finding out whether something is possible, not building it.
Lightweight feasibility / research mode. Spend ~10-30 minutes finding out whether something is possible, not building it.
The output is a findings document with citations, not a plan and not code.
/harness or /harness-plan)The PM Orchestrator dispatches only the Requirement Analyst (lightly), then YOU (the PM or the user) do the exploration directly. There is no Architect, no Gate, no Developer, no QA.
docs/tasks.md with mode: explore and stage: exploring.docs/features/<task-slug>/ and PM_LOG.md.01_REQUIREMENT_ANALYSIS.md containing:
pip show, npm info, curl-ing an API, etc.)findings.md in the task directory. Format:
# Findings: <question>
## Answer
<one-paragraph direct answer>
## Evidence
- Source 1: <link or file:line> — <what it shows>
- Source 2: ...
## Implications for our project
- <bullet>
- <bullet>
## Recommended next step
- <one of: "proceed via /harness-plan with approach X", "abandon — see evidence", "explore deeper — open follow-up task Y">
stage: explored..harness/insight-index.md (see .harness/rules/05-insight-index.md).docs/features/<task-slug>/
01_REQUIREMENT_ANALYSIS.md (light, ~1 page)
findings.md (the actual deliverable)
PM_LOG.md
No 02-07.
/harness-explore is ~15-20% of full pipeline cost. Don't bundle multiple unrelated questions — one explore task = one question.
/harness-plan or /harness.findings.md. The artifact IS the point of the task.npx claudepluginhub alan-ift/harness-kit --plugin harness-kitGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.