parallax: Instructions file for Claude Code

CLAUDE.md

parallax CLAUDE.md is an instructions file for Claude Code from sani-savaliya/parallax. It costs 744 tokens per session, scanned A, original, MIT.

Project instructions for Parallax, a tool that checks whether an output still looks correct after it is rendered. Rendered output is the visible result produced from code, such as an image or highlighted text.

In plain words
What is it for?
Working on the Parallax library, command-line tool, MCP server, documentation, tests, and evaluation code.
Why use it?
They guide agents through the project’s planned tests, design rules, and safeguards when changing the verification system.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md.

This is sani-savaliya/parallax's own configuration. It tells Claude Code how to work on parallax itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything parallax configures →

Reuse

Borrowing it

Nothing to install: this file belongs to sani-savaliya/parallax. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/sani-savaliya/parallax/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/sani-savaliya/parallax

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for parallax CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/sani-savaliya/parallax/claude-md/github.svg)](https://agentmods.dev/instructions/sani-savaliya/parallax/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/sani-savaliya/parallax/claude-md"><img src="https://agentmods.dev/badge/instructions/sani-savaliya/parallax/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for parallax CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/sani-savaliya/parallax/claude-md"><img src="https://agentmods.dev/badge/instructions/sani-savaliya/parallax/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 744 This file is loaded in full into every session.
When invoked 744 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00744 $0.00744
Opus 5 $0.00372 $0.00372
Sonnet 5 $0.00149 $0.00149
Haiku 4.5 $0.00074 $0.00074

Measured 9d ago against content hash 31b2c360daee, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

parallax CLAUDE.md 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 9d 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.

CLAUDE.md · 30 lines

How it starts

The opening of the file, as written. The whole thing — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Parallax — project instructions

Render-in-the-loop self-verification primitive for coding agents + a factorial-decomposition study of why it works. Part of the SunnyCode workspace.

Status

M1 (library MVP) usable. The crosscheck primitive runs end-to-end with two adapters (SQL-result-shape, regex-highlight), an offline heuristic critic + an Anthropic vision critic, a CLI (parallax), and an MCP server (parallax-mcp). The full design package lives in docs/; start from docs/TASKS.md (TDD-first, split into M1 library MVP and M2 study). Read docs/EVAL.md before touching any eval code and docs/RED-TEAM.md to understand why the design is shaped the way it is. M2 (the factorial study) is not started.

Names & handles

  • Brand: Parallax · PyPI/npm dist: parallax-verify · import package: parallax · repo: sani-savaliya/parallax
  • API function: crosscheck(artifact, render_fn) (kept deliberately — only the product name changed away from the collision-heavy "crosscheck").
  • Domains available if wanted: parallaxverify.com, parallax.sh.

Build & test (planned)

  • Language: Python. Package layout src/parallax/, tests/, pyproject.toml, ruff + mypy.
  • TDD is mandatory — tests before implementation, 80%+ coverage (unit + integration + the eval harness has its own fixtures).
  • Run tests from this project root (see docs/TASKS.md for exact commands once scaffolded).

Key design invariants (do not violate)

  • Single multimodal model across all model arms — never a text-LLM vs a separate VLM (that's a sixth confound).
  • The feature set F shown to model arms contains no verdict flags — only raw/derived quantities. All thresholding lives in the deterministic A5 arm, whose thresholds are disjoint from the defect-generating parameters. (This is the leak that collapsed v1.)
  • Compute is measured, not gated — never abort a run on token-equality (selection on the dependent variable). Match on cost + passes + FLOP proxies; report them.
  • Symmetrize tuning effort, never read-back outcome — capping the render to what the text reaches is mediator/collider adjustment that erases real effects.
  • Renders sandbox-safe: DuckDB with httpfs/COPY/external-file access disabled, timeouts, no network; beware the subprocess socket-monkeypatch bypass (see docs/ARCHITECTURE.md).

Read the full file on GitHub · 30 lines

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. 9d ago First seen · 30 lines · 744 tokens per session scan A 31b2c360daee

Subscribe to this mod's changes

parallax CLAUDE.md is an instructions file published in the GitHub repository sani-savaliya/parallax (0 stars, last pushed 1mo ago), licensed MIT. It adds 744 tokens to every session, about $0.0037 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 2026-08-31.

Related

Other instructions, from other repositories

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,153 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,469 tokens