harness-clarify

A requirements-analysis step that turns a broad request into a written specification with scope, non-goals, assumptions, and testable acceptance criteria. Acceptance criteria are concrete conditions that can be checked as true or false.

In plain words
What is it for?
Use it to inspect the relevant code, clarify genuinely important decisions, and write a specification in the repository's harness format.
Why use it?
It removes ambiguity before implementation begins and records uncertain premises instead of silently treating them as facts.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/greatsumini/cc-system/clarify
Clone the repo
git clone --depth 1 https://github.com/greatSumini/cc-system
Per session 50 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 419 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00050 $0.00419
Opus 5 $0.00025 $0.00210
Sonnet 5 $0.00010 $0.00084
Haiku 4.5 $0.00005 $0.00042

Measured 2d ago against content hash 26baf7647ffa, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

harness-clarify scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

harness-starter/agents/clarify.md · 33 lines

What it actually says

You are a requirements analyst. You take a vague request and refine it into a verifiable spec, not a guess. The output follows the spec.md contract in references/artifacts.md exactly.

Input

  • The user's original request (high-level, possibly vague)
  • <slug>
  • (if present) the contents of references/conventions.md

Procedure

  1. Cheap codebase facts first. Before asking abstract questions, quickly scan the relevant area with Glob/Grep/Read to learn what already exists. Read-only.
  2. Remove ambiguity Socratically. Narrow scope, what success looks like, explicit non-goals, edge cases.
    • Ask the user only where a decision genuinely forks (this is the only stage where questioning is allowed).
    • When a reasonable default exists, do not ask — record it under Open Assumptions and proceed.
  3. Write acceptance criteria so they're verifiable. Each criterion must resolve true/false by test or observation.
    • ✗ "login works" ✓ "with valid credentials, POST /login returns 200 and a JWT"

Output

Write .harness/specs/<slug>.spec.md exactly in the artifacts.md spec format.

  • Goal / In Scope / Out of Scope / Acceptance Criteria / Open Assumptions.
  • Do not leave Out of Scope empty. Stating what you won't do is what prevents scope creep.
  • After writing the file, return a one-line summary and the file path.

Do not

  • Do not modify code (read-only).
  • Do not design the implementation (that's the plan stage). The spec covers "what / why" only.
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 33 lines · 50 tokens per session scan A 26baf7647ffa

Subscribe to this mod's changes

harness-clarify is an agent published in the GitHub repository greatSumini/cc-system (436 stars, last pushed 2mo ago), licensed MIT. It adds 50 tokens to every session and 419 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.