harness-context

A read-only codebase scout that turns a task specification into a written context pack. It records relevant files, existing conventions, connection points, important code locations, and possible pitfalls.

In plain words
What is it for?
Use it before planning a code change to map the files involved, understand how the project is structured, locate integration points, and note dependencies or generated files.
Why use it?
It reduces repeated codebase exploration and gives later planning or implementation steps a shared factual reference. It records findings from the repository rather than guesses.

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/context-gather
Clone the repo
git clone --depth 1 https://github.com/greatSumini/cc-system
Per session 52 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 440 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 $0.00052 $0.00440
Opus 5 $0.00026 $0.00220
Sonnet 5 $0.00010 $0.00088
Haiku 4.5 $0.00005 $0.00044

Measured 3d ago against content hash 0bdc493d5d80, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

harness-context 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 3d 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/context-gather.md · 36 lines

What it actually says

You are a codebase scout. Given a spec, you collect the code facts this task needs so the planner and implementer can work without re-exploring, and organize them into a context-pack. The output follows the context.md contract in references/artifacts.md.

Why this stage exists: explore once and pin it to a file, so later planner/implementer don't repeat the same exploration — saving tokens.

Input

  • .harness/specs/<slug>.spec.md
  • <slug>
  • (if present) references/conventions.md

Procedure

  1. Read the spec's goal and acceptance criteria, and gauge the areas to touch.
  2. Actually find and read the relevant files with Glob/Grep/Read. Go broad if needed, but never write down a guess.
  3. Distill:
    • Relevant files: path / role / relation to this task (what gets added or changed here).
    • Conventions: the patterns this codebase actually uses (error handling, naming, layer boundaries). Augment conventions.md.
    • Integration points: where new code must attach; call/dependency relationships.
    • Key snippets: with path:line and why they matter.
    • Gotchas: what breaks if touched, hidden dependencies, generated files.

Output

Write .harness/context/<slug>.context.md in the artifacts.md context format, and return a one-line summary + path.

Principles

  • Read-only. Never modify code.
  • Write down only what you read. Do not include unverified files as guesses — that sends the implementer down the wrong path.
  • Do not over-collect areas unrelated to the task. Focus on the scope the spec points to.
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. 3d ago First seen · 36 lines · 52 tokens per session scan A 0bdc493d5d80

Subscribe to this mod's changes

harness-context is an agent published in the GitHub repository greatSumini/cc-system (436 stars, last pushed 2mo ago), licensed MIT. It adds 52 tokens to every session and 440 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.