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/mrwogu/promptscript/claude-mdgit clone --depth 1 https://github.com/mrwogu/promptscriptWrote 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/mrwogu/promptscript/claude-md)<a href="https://agentmods.dev/instructions/mrwogu/promptscript/claude-md"><img src="https://agentmods.dev/badge/instructions/mrwogu/promptscript/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 | $0.02477 | $0.02477 |
| Opus 5 | $0.01239 | $0.01239 |
| Sonnet 5 | $0.00495 | $0.00495 |
| Haiku 4.5 | $0.00248 | $0.00248 |
Grade A, and why
promptscript 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 4d 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.
This is a copy
83% identical to promptscript copilot-instructions.md — 216 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 298 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Project
You are an expert TypeScript developer working on PromptScript - a language and toolchain for standardizing AI instructions across enterprise organizations.
PromptScript compiles .prs files to native formats for GitHub Copilot,
Claude Code, Cursor, and other AI tools.
You write clean, type-safe, and well-tested code following strict TypeScript practices.
Tech Stack
typescript, Node.js 20+, Nx + pnpm
Architecture
The project is organized as a monorepo with these packages:
flowchart TB
subgraph packages
core[core - Types, errors, utilities]
parser[parser - Chevrotain-based parser]
resolver[resolver - Inheritance & import resolution]
validator[validator - AST validation rules]
compiler[compiler - Pipeline orchestration]
formatters[formatters - Output formatters]
cli[cli - Command-line interface]
end
cli --> compiler
cli --> resolver
compiler --> resolver
compiler --> validator
compiler --> formatters
resolver --> parser
parser --> core
resolver --> core
validator --> core
formatters --> core
Context
Key Libraries
-
Parser: Chevrotain
-
CLI: Commander.js
-
Testing: Vitest
-
Linting: ESLint + Prettier
-
Project: PromptScript
Engineering Standards
- Use code-review-graph MCP tools BEFORE Grep/Glob/Read when exploring the codebase
- Exploring code: use semantic_search_nodes or query_graph instead of Grep
- Understanding impact: use get_impact_radius instead of manually tracing imports
- Code review: use detect_changes + get_review_context instead of reading entire files
- Finding relationships: use query_graph with callers_of/callees_of/imports_of/tests_for
- Architecture questions: use get_architecture_overview + list_communities
- Fall back to Grep/Glob/Read only when the graph does not cover what you need
- Strict mode enabled
- Never use
anytype - useunknownwith type guards - Use
unknownwith type guards instead of any - Prefer
interfacefor object shapes - Use
typefor unions and intersections - Named exports only, no default exports
- Explicit return types on public functions
- Files:
kebab-case.ts - Classes:
PascalCase - Interfaces:
PascalCase - Functions:
camelCase - Variables:
camelCase - Constants:
UPPER_SNAKE_CASE - Use custom error classes extending
PSError - Always include location information
- Provide actionable error messages
- Test files:
*.spec.tsnext to source - Follow AAA (Arrange, Act, Assert) pattern
- Framework: Vitest
- Target >90% coverage for libraries
- Use fixtures for parser tests
- When refactoring formatter section methods (e.g. splitting context() into project() + techStack() + architecture()), add a test verifying that ALL input block content still appears in the output — not just the newly extracted subsections
- Golden files are snapshots of correct behavior, not correct by definition — before regenerating golden files, verify the diff represents an intentional change, not a regression that golden files would lock in
- keepInSync: when adding or changing block keywords (e.g. @knowledge, @guards), always update ALL THREE syntax highlighters: (1) Pygments lexer: docs_extensions/promptscript_lexer.py, (2) VS Code TextMate grammar: apps/vscode/syntaxes/promptscript.tmLanguage.json, (3) Playground Monaco language: packages/playground/src/utils/prs-language.ts
- branchStrategy: gitflow — branch from main, never commit to main
- newTask: follow the new-task workflow: branch, atomic commits, full verification pipeline, push, PR, watch CI
- prMonitoring: use
gh pr checks --watchto monitor CI status; do not consider work done until all checks pass - verification: follow the verify workflow after any code change - all eight steps, in order
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.
- 4d ago First seen · 298 lines · 2,477 tokens per session scan A 5aeac0293652
promptscript CLAUDE.md is an instructions file published in the GitHub repository mrwogu/promptscript (391 stars, last pushed 3d ago), licensed MIT. It adds 2,477 tokens to every session, about $0.0124 per session on Opus 5. A static security scan graded it A with 0 findings. It is 83% identical to promptscript copilot-instructions.md, differing in 216 lines, and is treated as a copy.
Other instructions, from other repositories
ai-prompts AGENTS.md
Instructions for liatrio-labs/ai-prompts, covering ai agent instructions, context marker, local skill scripts, standard workflow and initialize.
gortex CLAUDE.md
Claude Code instructions for zzet/gortex, covering gortex, build & test, codebase overview, discovery (read once, then keep using) and llm provider (powers ask and searchsymbols assist: modes).
agent-toolkit AGENTS.md
Instructions for ulises-jeremias/agent-toolkit, covering agents.md — ai agent contract, what this toolkit does, repository structure, operating rules and how to add a skill.
core-ai CLAUDE.md
Instructions for helius-labs/core-ai, covering core ai, compiler-managed sync, mcp server pin, skill versioning and router surface maintenance.
ChatCrystal AGENTS.md
AGENTS.md instructions for ZengLiangYi/ChatCrystal, covering agents.md, project overview, commands, development (server port 3721 + client port 13721) and build for production.
core-ai AGENTS.md
Instructions for helius-labs/core-ai, covering core ai — agent instructions, repository overview, mcp server setup, api key setup and public tool surface.