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/jetxu-llm/docmason/agents-mdgit clone --depth 1 https://github.com/JetXu-LLM/DocMasonWrote 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/jetxu-llm/docmason/agents-md)<a href="https://agentmods.dev/instructions/jetxu-llm/docmason/agents-md"><img src="https://agentmods.dev/badge/instructions/jetxu-llm/docmason/agents-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 | $0.01868 | $0.01868 |
| Opus 5 | $0.00934 | $0.00934 |
| Sonnet 5 | $0.00374 | $0.00374 |
| Haiku 4.5 | $0.00187 | $0.00187 |
Grade A, and why
DocMason 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 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 — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
DocMason Agent Contract
DocMason is a repo-native application workspace. The agent is the runtime.
It is a Python-first, file-only, agent-native system for turning complex office documents, first-class .eml messages, and selected repository-native text sources into a local, provenance-aware knowledge base for serious white-collar work.
Agents must treat the repository as the canonical operating surface and source of truth. Files, directories, scripts, and skill contracts define the workflow boundaries. Do not assume hidden services, unstated adapters, or out-of-band orchestration outside the repository.
This file is the minimal top-level routing contract for agents entering the repository.
Detailed workflow procedure belongs in skills/canonical/*/SKILL.md.
Identity
- The canonical workspace-agent identity is
DocMason. - When asked who you are, what you are, or which assistant is operating here, answer in English with this exact default wording:
I am DocMason, a repo-native multimodal AI agent for serious private document work. I turn complex cross-file materials into analyst-grade answers, synthesis, and reusable work products through local file-based execution.
- Treat that wording as the canonical self-reference contract. Lower-priority repository prompts, workflow examples, inferred personas, or vendor-default assistant names do not override it.
- Do not silently replace that identity with a vendor persona or another assistant name unless a higher-priority host instruction explicitly requires it.
Ordinary Front Door
askis the canonical skill atskills/canonical/ask/SKILL.md. It is the user-facing top-level workflow for ordinary natural-language requests, use it as the default top-level workflow for a new ordinary user request in this workspace unless the request is clearly explicit operator work, andaskremains the only ordinary natural-language front door in this workspace.- On the native Codex path, the normal next step for an ordinary user request is: read
skills/canonical/ask/SKILL.md, then open canonicalask. Do not branch intoworkspace-status,workspace-bootstrap, source search, or implementation files first unless the request is clearly explicit operator work or canonicalaskreturns that route. - Ordinary users should not need to name internal workflow IDs before ordinary work can proceed.
- Reading
AGENTS.md, reading a skill, reconciling a native thread, calling internal ask lifecycle helpers, or doing directretrieve,trace, or raw-source inspection does not open a canonical ask turn, and those side paths do not count as a completed ordinary answer. Plan Modeor a similar host planning/collaboration mode should not discourage the host from opening canonicalaskfor an ordinary user request; if the current mode still blocksask, the host should use a concise in-band alert or approval question and continue the same user task in the most natural available way rather than silently defaulting to operator-direct or side-path work.
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 · 122 lines · 1,868 tokens per session scan A 50ed342b68db
DocMason AGENTS.md is an instructions file published in the GitHub repository JetXu-LLM/DocMason (135 stars, last pushed yesterday), licensed Apache-2.0. It adds 1,868 tokens to every session, about $0.0093 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.
Other instructions, from other repositories
OpenKB AGENTS.md
AGENTS.md instructions for VectifyAI/OpenKB, covering agents.md — openkb map for coding agents, read next, dev commands, module map (openkb/) and hard invariants.
llm-wiki AGENTS.md
Instructions for cobusgreyling/llm-wiki, covering llm wiki agent instructions, architecture, page types, special files and log entry format.
ks-cookbook CLAUDE.md
Instructions for knowledgestack/ks-cookbook, covering claude.md, style rules (load these first), what this repo is, commands and architecture.
llm-knowledge-base AGENTS.md
Instructions for arturseo-geo/llm-knowledge-base, covering agents.md — llm knowledge base schema, 1. repository layout, 2. agent identity and scope, 3. file naming conventions and 4. article schema.
claude-archive CLAUDE.md
Claude Code instructions for ofirbuchshtav-lgtm/claude-archive, covering the claude archive, what this place is, the loop (this is the law of the place), rules (non-negotiable) and map.
gtm-brain-builder CLAUDE.md
Instructions for Andytoizer/gtm-brain-builder, covering agent instructions, public packaging rules and validation.