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.
npx agentmods add instructions/vinicq/falsegreen-skill/claude-mdgit clone --depth 1 https://github.com/vinicq/falsegreen-skillWrote 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.
[](https://agentmods.dev/instructions/vinicq/falsegreen-skill/claude-md)<a href="https://agentmods.dev/instructions/vinicq/falsegreen-skill/claude-md"><img src="https://agentmods.dev/badge/instructions/vinicq/falsegreen-skill/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00583 | $0.00583 |
| Opus 5 | $0.00292 | $0.00292 |
| Sonnet 5 | $0.00117 | $0.00117 |
| Haiku 4.5 | $0.00058 | $0.00058 |
Grade A, and why
falsegreen-skill 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md - falsegreen-skill
Project context for Claude Code and other agents opening this repository.
What This Project Is
falsegreen-skill is a public LLM skill for finding false-positive tests:
tests that stay green even when the implementation is wrong.
It is not Claude-specific. The same protocol is packaged for Claude Code, Codex, Gemini, Cursor, plain LLM prompts, API usage, and the npm CLI.
The companion falsegreen project is the deterministic Python scanner. This
repository is the semantic, multi-host skill layer.
Public Repo Boundary
This repo should contain only public product assets:
SKILL.md- canonical J1-J6 protocol and structural catalog rules.reference.md- per-language pattern catalog and look-alike exemptions.schema/- canonical machine-readable output contracts.skills/- plugin skill entry points..claude-plugin/,.codex-plugin/,.antigravity-plugin/,.agents/,.gemini/- host packaging metadata (.antigravity-plugin/= Antigravity CLI plugin foragy plugin install;.agents/skills/= Antigravity CLI workspace skill;.gemini/= legacy Gemini CLI copy).contexts/- host-specific usage guides.bin/andscripts/- CLI and packaging/validation scripts.examples/- public examples only.
Do not add private datasets, adjudication notes, unpublished paper drafts, internal agent outputs, or benchmark artifacts here.
Maintenance Rules
- Keep
SKILL.md,llm.md,GEMINI.md, and host contexts consistent when the protocol changes. - Treat
schema/finding.jsonandschema/report.jsonas the source of truth for JSON output. - Supported languages in public docs: Python, TypeScript, JavaScript, and Robot
Framework (the catalog covers all four, mirroring the static scanners). Gherkin
and Tavern are covered as secondary semantic passes in
reference.md. - The skill is a superset of the three static scanners: every structural code in
falsegreen / falsegreen-js / robotframework-falsegreen must appear in
reference.md, plus the AI-only semantic codes (S-series). - Prefer thin host entry points that reference the canonical protocol instead of duplicating the full catalog.
- Run
npm run validatebefore release. - Run
npm run build:targetswhen preparing standalone Claude/Gemini skill packages.
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.
- 5d ago First seen · 63 lines · 583 tokens per session scan A 309546ed593e
falsegreen-skill CLAUDE.md is an instructions file published in the GitHub repository vinicq/falsegreen-skill (2 stars, last pushed 4d ago), licensed MIT. It adds 583 tokens to every session, about $0.0029 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.
Other instructions, from other repositories
wyx CLAUDE.md
Instructions for jlifyio/wyx, covering claude.md, project overview, architecture, skills and hooks.
cli copilot-instructions.md
Copilot instructions for reporails/cli, covering dashboard app, commands, architecture and constraints.
cli python.instructions.md
Instructions for reporails/cli, a project described as: AI Instructions Diagnostics | Supported agents: Claude | Codex | Copilot | Cursor | Gemini.
rules copilot-instructions.md
Copilot instructions for reporails/rules, covering reporails rules framework, tech stack, test commands, build test image (first time or after runtime/ changes) and run all rules.
rules CLAUDE.md
Claude Code instructions for reporails/rules, covering reporails framework, tech stack, terminology, file discovery and initialization.
ai-dial-core CLAUDE.md
Claude Code instructions for epam/ai-dial-core, covering claude.md, build & run, set credentials via environment variables, build (skip tests) and run all tests.