What is doc-agent AGENTS.md?

A set of project instructions for doc-agent, a command-line tool that extracts structured information from PDFs and images and searches documents by meaning. It explains the project's setup, code layout, tools, and required workflow.

Use it when working on doc-agent features, fixes, setup, document extraction, semantic search, or its Model Context Protocol integration.

It gives a coding agent the project-specific rules and context needed to change doc-agent safely and run the right checks.

doc-agent AGENTS.md — Instructions file by prosdevlab

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 instructions/prosdevlab/doc-agent/agents-md

Clone the repo

git clone --depth 1 https://github.com/prosdevlab/doc-agent

Made for: Codex, OpenCode.

Per session 1,552 This file is loaded in full into every session.

When invoked 1,552 The same file — it is already loaded in full.

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.

Model	Per session	Once invoked
Fable 5	$0.01552	$0.01552
Opus 5	$0.00776	$0.00776
Sonnet 5	$0.00310	$0.00310
Haiku 4.5	$0.00155	$0.00155

Measured yesterday against content hash 2f15ece659af, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

doc-agent AGENTS.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 yesterday.

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.

AGENTS.md · 181 lines

How it starts

The opening of the file, as written. The whole thing — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

This is the developer documentation for the doc-agent project.

CRITICAL: When starting any new task, feature, or fix, you MUST follow The Drill™ outlined in WORKFLOW.md.

Project Overview

doc-agent is a document extraction and semantic search CLI with MCP (Model Context Protocol) integration. It allows extracting structured data from PDFs and images (e.g., invoices, receipts, bank statements) using Vision AI and performing semantic search over indexed documents.

Tech Stack:

Runtime: Node.js >= 22 (LTS)
Package Manager: pnpm (using workspaces)
Language: TypeScript
Build System: Turborepo + tsup
Linter/Formatter: Biome
Testing: Vitest + ink-testing-library
AI Providers: Ollama (default, local), Google Gemini (cloud)
Database: SQLite via better-sqlite3 + Drizzle ORM
Vector Database: LanceDB (via vectordb package)
CLI Framework: Ink (React for CLIs) + Commander.js
OCR: Tesseract.js (WASM-based, fully local)

Repository Structure

The project is organized as a monorepo using pnpm workspaces:

packages/
├── cli/           # CLI entry point and MCP server
├── core/          # Shared types and interfaces
├── extract/       # Document extraction (Gemini, Ollama)
├── storage/       # SQLite persistence (Drizzle ORM)
└── vector-store/  # Vector database for semantic search

Setup Commands

# Install dependencies
pnpm install

# Build all packages
pnpm build

# Run CLI locally during development
pnpm dev extract examples/invoice.pdf

# Run MCP server locally
pnpm mcp

Development Workflow

See WORKFLOW.md for the complete development process, including:

Finding work
Planning
Implementation
Quality Checks
Committing & PRs

Working on the CLI

The CLI is the main entry point (packages/cli), built with Ink (React for CLIs).

Run CLI: pnpm --filter @doc-agent/cli dev ... (or just pnpm dev from root)
Command Structure:
- doc extract <file>: Extract data from a document.
- doc mcp: Start the MCP server.
- doc search <query>: (Coming soon) Search indexed documents.

Read the full file on GitHub · 181 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.

yesterday First seen · 181 lines · 1,552 tokens per session scan A 2f15ece659af

Subscribe to this mod's changes

doc-agent AGENTS.md is an instructions file published in the GitHub repository prosdevlab/doc-agent (0 stars, last pushed 8mo ago), licensed MIT. It adds 1,552 tokens to every session, about $0.0078 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.

Data and AI Documents Research

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 · 5,182 tokens

buildNext

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

doc-agent AGENTS.md